Bao Xiaobin, Yang Yingna, Qi Jianqing, He Yanbing
School of Architecture and Art Design, Henan Polytechnic University, Jiaozuo, Henan, China.
School of Marxism, Henan Polytechnic University, Jiaozuo, Henan, China.
PLoS One. 2025 Aug 1;20(8):e0329356. doi: 10.1371/journal.pone.0329356. eCollection 2025.
By revealing the spatial distribution characteristics and driving mechanisms of 1,027 traditional Chinese villages in the Loess Plateau, as announced in six batches up to 2023, this article provides a theoretical basis for formulating scientific and differentiated protection and development strategies. Utilizing the ArcGIS 10.8 platform and the GeoDetector model, this study comprehensively applies methods including the nearest neighbor index, standard deviation ellipse, kernel density estimation, and spatial autocorrelation to systematically analyze the spatial pattern of traditional villages, and quantitatively reveals their key driving factors and interactions through the GeoDetector. The results show: (1) Traditional villages in the Loess Plateau present a significant clustered spatial distribution (Nearest Neighbor Index R = 0.47, Moran's I = 0.189), with an overall pattern of "dense in the east and west, sparse in the center," primarily concentrated within an elliptical area with an flattening ratio of 0.664 and a directional angle of 84.5°, and with high-density areas located in Shanxi Province, the Yellow River basin along the Shanxi-Shaanxi border, and the Hehuang Valley. (2) Spatially, five major core clusters are formed, presenting a layered structure of "core clustering, gradient transition," and exhibiting a non-stable diffusion trend towards the periphery. (3) The spatial differentiation is a result of the synergistic driving of human and natural factors, with human factors having a stronger overall explanatory power. Among them, the density of cultural heritage sites (q = 0.546) is the primary driving factor, followed by annual precipitation (q = 0.458), population size (q = 0.457), and urbanization rate (q = 0.354). The interaction between factors is significant and mostly shows non-linear enhancement, for instance, the interactive explanatory power of population size and the density of cultural heritage sites is as high as 0.852.
通过揭示截至2023年分六批公布的黄土高原1027个中国传统村落的空间分布特征及驱动机制，本文为制定科学、差异化的保护与发展策略提供了理论依据。本研究利用ArcGIS 10.8平台和地理探测器模型，综合运用最近邻指数、标准差椭圆、核密度估计和空间自相关等方法，系统分析传统村落的空间格局，并通过地理探测器定量揭示其关键驱动因素及相互作用。结果表明：（1）黄土高原传统村落呈现显著的集聚型空间分布（最近邻指数R = 0.47，莫兰指数I = 0.189），总体格局为“东西部密集，中部稀疏”，主要集中在扁率为0.664、方向角为84.5°的椭圆区域内，高密度区域位于山西省、晋陕边界黄河流域以及河湟谷地。（2）在空间上形成了五个主要核心集群，呈现出“核心集聚、梯度过渡”的分层结构，并向周边呈现出非稳定扩散趋势。（3）空间分异是人类和自然因素协同驱动的结果，其中人类因素总体解释力更强。其中，文化遗产地密度（q = 0.546）是首要驱动因素，其次是年降水量（q = 0.458）、人口规模（q = 0.457）和城镇化率（q = 0.354）。各因素间的相互作用显著，且大多呈现非线性增强，例如人口规模与文化遗产地密度的交互解释力高达0.852。
