Functional traits and climate drive interspecific differences in disturbance-induced tree mortality.

With climate change, natural disturbances such as storm or fire are reshuffled, inducing pervasive shifts in forest dynamics. To predict how it will impact forest structure and composition, it is crucial to understand how tree species differ in their sensitivity to disturbances. In this study, we investigated how functional traits and species mean climate affect their sensitivity to disturbances while controlling for tree size and stand structure. With data on 130594 trees located on 7617 plots that were disturbed by storm, fire, snow, biotic or other disturbances from the French, Spanish and Finnish National Forest Inventory, we modeled annual mortality probability for 40 European tree species as a function of tree size, dominance status, disturbance type and intensity. We tested the correlation of our estimated species probability of disturbance-mortality with their traits and their mean climate niches. We found that different trait combinations controlled species sensitivity to disturbances. Storm-sensitive species had a high height-dbh ratio, low wood density and high maximum growth, while fire-sensitive species had low bark thickness and high P50. Species from warmer and drier climates, where fires are more frequent, were more resistant to fire. The ranking in disturbance sensitivity between species was overall consistent across disturbance types. Productive conifer species were the most disturbance-sensitive, while Mediterranean oaks were the least disturbance-sensitive. Our study identified key relations between species functional traits and disturbance sensitivity, that allows more reliable predictions of how changing climate and disturbance regimes will impact future forest structure and species composition at large spatial scales.