Developing a point process model for ecological risk assessment of pine wilt disease at multiple scales

Abstract

Pine wilt disease (PWD) is one of the most serious conifer diseases, with further expansion expected under ongoing climate change. Sustainable forest management and effective pest control of pine forests therefore requires rigorous exploration of potential PWD risk areas under current and future climate conditions. To predict potential PWD risk areas in Japan, we constructed an inhomogeneous Poisson point process (IPP) model, which allows the use of strongly biased data. We examined both current and near-future climatic conditions (2026–2050, five general circulation model under two representative concentration pathways [RCPs]). Occurrence data were obtained on three different spatial scales (national, regional and local) along with eight bioclimatic variables. The resultant model was able to correct the data bias caused by using these three different data sources and showed high predictive power as follows: (1) potential risk areas will increase more than the current PWD distribution; (2) mean annual temperature will have the highest effect of the eight predictor variables; (3) high-risk areas will expand northwards and/or upwards in the near future, at a maximum of ca. +58.6% under RCP 8.5 and a minimum of ca. +15.9% under RCP 2.6 and low risk areas will decrease. The high-resolution PWD risk maps created with the IPP model will therefore aid future pest control and forest management strategies at local, regional, and national scales.