Biological invasions Erode China's forest carbon stocks: Spatiotemporal expansion and edge preferences of Pine Wilt Disease.

Pine Wilt Disease (PWD), a devastating epidemic, poses a severe threat to the carbon sink function and stability of China's forest ecosystems. However, accurate long-term quantification of carbon loss remains challenging due to historical data gaps and scale mismatches between macro-statistics and micro-ecological processes. To accurately quantify the forest carbon loss caused by PWD, this study integrates multi-source remote sensing products with national monitoring data across China. Methodologically, we proposed an innovative ensemble learning-based "stand-to-tree" scale conversion strategy to reconstruct mean biomass per tree (MBT) from macroscopic remote sensing observations, and designed a bottom-up approach to aggregate carbon loss from local outbreak patches to the national level. Results indicate that: (1) From 1998 to 2022, PWD caused a cumulative carbon loss of 12.06 Tg C (95% scenario envelope: 9.88-15.79 Tg C). The spatiotemporal evolution exhibited a "latent accumulation-diffusion-explosive growth" pattern, with 2017 as a critical tipping point, where hotspots shifted from the southeast coast to the Yangtze River basin and northern suitable habitats. (2) Landscape analysis revealed that PWD outbreaks were predominantly distributed near forest edges and were spatiotemporally associated with subsequent forest-boundary retreat. This study provides a detailed national carbon loss inventory and a scientific basis for understanding the spatial association between disease outbreaks and landscape patterns, facilitating the formulation of ecosystem-based control strategies.