Revealing the spatial heterogeneity of pine wilt disease spread in China through Geographical Pattern Causality analysis.

Abstract

Pine wilt disease (PWD) is one of the most destructive invasive forest diseases in China, yet the environmental drivers underlying its spatially heterogeneous spread remain poorly understood. This study aimed to characterize the spatial and temporal patterns of natural PWD dispersal, identify persistent hotspots, and determine region-specific environmental drivers of PWD habitat suitability. We developed an integrated framework combining dispersal screening, biomod2 ensemble modelling, SaTScan spatiotemporal scan statistics, and Geographical Pattern Causality (GPC) analysis to examine PWD spread and its environmental drivers at national and regional scales. Natural dispersal increased markedly after 2016 and reached a peak of 9728 newly affected subcompartments in 2020. At the national scale, Altitude, mean diurnal temperature range (Bio2), and normalized difference vegetation index (NDVI) were the dominant determinants of habitat suitability, with relative importance values of 27.52%, 25.29%, and 21.87%, respectively. However, GPC analysis confirmed pronounced spatial heterogeneity among the SaTScan-identified hotspot regions during the outbreak period. In hotspot region A, Aspect and Slope showed the strongest positive causal relationships on PWD suitability, with causal strengths of 0.361 and 0.284, respectively. In contrast, Altitude and Slope showed strong negative effects in hotspot region C, with strengths of 0.591 and 0.520, respectively. Wind speed showed negative causal relationships with PWD habitat suitability across all hotspot regions. These findings demonstrate that the environmental drivers of PWD vary substantially across spatial scales and hotspot regions, highlighting the importance of incorporating regional heterogeneity into PWD monitoring, risk assessment, and differentiated management.