Historical corridor proximity and the spatial differentiation of traditional villages along the Xiaohe Ancient Road, China

Traditional villages in multi-ethnic mountainous regions are often described as following historical transport corridors, but testable relationships among corridor proximity, environmental conditions, and sociocultural differentiation are lacking. This study analyzes 122 officially listed traditional villages along the Xiaohe Ancient Road, Guangxi, combining village inventories, geospatial data, village-level ethnic records, and gazetteers. Point-pattern, spatial autocorrelation, negative binomial and multinomial logistic regression, and GeoDetector analyses were applied with threshold and grid-scale robustness checks. The villages are significantly clustered (nearest-neighbor ratio = 0.658) and positively autocorrelated at the grid scale (Moran’s I = 0.154, both p < 0.001), forming multi-core hotspots with corridor-oriented extension and concentrating at low to mid elevations, on gentle slopes, and near rivers. Controlling for rivers, modern roads, and elevation, distance to the reconstructed main-route alignment remained negatively associated with grid-level village counts (B = −0.036, IRR = 0.965, p < 0.001). Village-level ethnic composition yielded only a category-specific, conditional, and exploratory association: main-route distance separated the moderately dominant type from the mixed type but not the highly dominant type, and was not corroborated by the GeoDetector ethnic extension. The cross-sectional design precludes causal inference. The findings support a shift from single-village protection toward corridor-based zoning and coordinated governance.