Holm oak death is accelerated but not sudden and expresses drought legacies.

The increase in abiotic and biotic stress driven by global change threatens forest ecosystems and challenges understanding of mechanisms producing mortality. Phytophthora spp. like P. cinnamomi (PHYCI) are among the most lethal pathogens for many woody species including Quercus spp. Dynamics of biotic agents and their hosts are complex and influenced by climatic conditions. We analysed radial growth trends of dead and live adult Quercus ilex trees from agrosilvopastoral open woodlands under intense land-use. A pronounced warming trend since the 1980s has coincided in these woodlands with high oak mortality rates generally attributed to PHYCI. Yet, tree mortality and latency of the pathogen could be expressed at variable time spans, whereas, like in many other forests worldwide, tree death could also be explained by other factors like drought. PHYCI was isolated from roots of all dead oaks from one region. Trees were younger than generally believed and ages of dead trees ranged between 38 and 230 years. Growth of dead trees reached a tipping point in 1980 and 1990 coincident with two-year extraordinary droughts. These dates set the start of growth declines up to 30 years before tree death. Live trees did not exhibit any recent growth decline. Tree growth was highly sensitive to climatic variability associated with water stress and climate-growth relationships suggested phenological changes since the 1980s. Live and dead trees showed differences in their sensitivity to moisture availability and temperature. The sensitivity of growth to climate was partially related to site environmental conditions. Simulated gross and net primary productivity were higher in live sites with less atmospheric demand for water. Tree death was not sudden but a slow multiannual process as expressed by radial growth declines likely triggered by drought. Regardless of the causal agent or mechanism, the observed mortality affected trees exhibiting negative drought and land-use legacies.