• Drought-driven forest mortality signals strengthen with increasing spatial scale. • Lagged climatic water balance effects reveal multi-year drought legacies. • Joint mean–dispersion models capture heterogeneity in mortality responses. • Scale-aware framework improves drought mortality prediction and monitoring. Drought-induced forest mortality has emerged as one of the most visible ecological responses to climate change, yet its predictability remains limited by the spatial scale of analysis. Here, we examine how climatic water balance predicts Scots pine ( Pinus sylvestris L.) mortality across three hierarchical levels of forest management in Poland, representing local, intermediate, and regional spatial scales. Using long-term mortality records and climatic data, we demonstrate that the strength of the drought–mortality relationship increases with spatial aggregation, as local stochasticity diminishes and climatic signals become more coherent. Our findings reveal that drought legacy effects extend over multiple years and that spatial heterogeneity, often treated as statistical noise, contains meaningful ecological information about forest vulnerability. These results highlight that identifying the appropriate spatial scale is crucial for improving mortality predictions and for developing adaptive management strategies aimed at enhancing forest resilience under ongoing climate change.
Tymińska-Czabańska et al. (Thu,) studied this question.