Unveiling the spatiotemporal dynamics and severity patterns of pine wilt disease in China: insights from fuzzy logic and data mining

Abstract

Pine wilt disease (PWD), a highly destructive invasive forest pest, has posed a significant threat to forest ecosystems in China since its introduction in 1982. While previous studies have focused primarily on large-scale patterns or specific regional cases, limited attention has been given to the regional dynamics and fine-scale drivers of PWD outbreaks. To address this gap, this study integrates fuzzy logic and data mining techniques to analyze epidemic sub-compartment data—the smallest spatial unit of PWD monitoring—based on nationwide records collected by the Biological Disaster Prevention and Control Center of the National Forestry and Grassland Administration from 2015 to 2023 in China. Our results reveal a clear temporal shift in PWD prevalence toward warmer, more humid regions, especially in ecologically fragile forests such as planted plantations. Fuzzy clustering analysis further highlight spatial heterogeneity in disease severity, and association rule mining uncovers distinct province-specific drivers. For example, climatic suitability emerged as the main driver in Guangdong, while topographic barriers and human activity were more influential in Jiangxi and Zhejiang, respectively. These findings offer a detailed understanding of both the temporal progression and regional complexity of PWD outbreaks, providing valuable insights to inform targeted control strategies, early warning systems, and precision forest management efforts.