Predicting drought-induced impacts on forests: mechanisms, challenges and novel approaches

Hotter and drier conditions under ongoing climate change are altering forest dynamics worldwide, frequently inducing events of forest mortality. These mortality events have profound implications for the carbon and water balances of ecosystems, from the plot to the global scale. However, their prediction remains a substantial research challenge, even if the mechanisms involved are relatively well understood. I will discuss this apparent paradox highlighting the limitations of current approaches, which rely almost exclusively on the use of traits and process-based modelling. These are powerful approaches but also face important challenges due to the scale- and context-dependency of trait effects on individual performance and vegetation dynamics. I will argue that better predictions require multi-scale approaches that complement the essentially reductionistic view based on traits with systems thinking. Current developments of forest monitoring schemes through continuous, on-site measurements and enhanced remote sensing tools provide excellent opportunities to reconcile our detailed mechanistic understanding at fine scales (e.g., ecophysiology) with patterns and processes at coarser scales.