Disentangling growth, water-use and nutritional constraints underlying forest dieback in drought-prone pine forests

Increasing drought frequency and intensity under climate change is driving widespread forest dieback and tree mortality, particularly in water-limited regions such as the Mediterranean Basin. Despite extensive research on drought-induced growth decline, mechanistic understanding of how interacting hydraulic, carbon, and nutritional constraints predispose individual trees to dieback remains incomplete. Most studies have focused on single processes, limiting our ability to identify early-warning signals and robust predictors of mortality risk. Here, we apply a multi-proxy, tree-level approach that links growth dynamics, water-use patterns, and nutrient status to diagnose drought-induced canopy dieback in Pinus sylvestris, Pinus pinaster, and Pinus halepensis forests along an aridity gradient in north eastern Spain. Within each stand, dominant and co-dominant trees were selected and classified as non-declining or declining trees based on crown defoliation. For each individual, we combined dendrochronological analyses with foliar elemental and isotopic composition, morphological traits, and soil properties measurements. Growth vulnerability to drought was quantified by combining long-term growth trajectories and growth–climate relationships. Leaf carbon and oxygen isotopic composition data (δ¹³C, δ¹⁸O) were used to infer intrinsic water-use efficiency (iWUE) and time-integrated stomatal conductance. Foliar macro- and micronutrient contents, expressed per unit leaf area, were measured to evaluate nutrient imbalances associated with drought stress and senescence, and were interpreted in relation to soil pH and nutrient availability. Tree size and needle morphological traits, including leaf mass per area (LMA), were also measured. We applied multivariate analyses (principal component analysis, partial least squares regression) to integrate physiological, nutritional, and structural variables, discriminate between non-declining and declining trees, and identify key predictors of crown defoliation as a proxy for vigour decline. Our approach provided mechanistic insight into how chronic drought stress propagates through coordinated reductions in growth and stomatal conductance combined with nutrient imbalances, while revealing species-specific pathways to dieback. Overall, this study addresses key gaps in drought-induced forest mortality research and contributes to improving diagnostic and prognostic frameworks on forest dieback under ongoing climate change.