Xylem traits and growth phenology predict growth and mortality response to defoliation in temperate forests

Defoliation outbreaks are biological disturbances that alter tree growth and mortality in temperate forests. Trees respond to defoliation in many ways; some recover rapidly, while others decline gradually or die. These differences may arise from species functional traits that constrain growth such as xylem anatomy, growth phenology or non-structural carbohydrate (NSC) storage, but this has not been shown. Although many studies address these phenomena, varied and idiosyncratic measures limit our ability to generalize and predict defoliation responses across species. I synthesized and translated published growth and mortality data into consistent standardized variables suitable for numerical models. I analyzed data from 32 studies, including 16 tree species and 10 defoliator systems from North America and Eurasia, and quantitatively compared responses to defoliation among species and tree functional groups using linear mixed-effects models.\n\nRelative growth decreased linearly or curvilinearly as defoliation stress accumulated across species. Growth decreased by only 5-20% following 100% defoliation in ring-porous Quercus, whereas growth of diffuse-porous hardwoods and conifers declined by 50-100%. Mortality increased exponentially with defoliation, more rapidly for Pinus and diffuse-porous species than for Quercus and Abies. Species-specific mixed models were best (R2c = 0.83-0.94), yet functional-group models lost little in terms of goodness-of-fit (R2c = 0.72-0.92), providing useful alternatives when species data is lacking. These responses are consistent with functional differences in wood growth phenology and NSC storage. Ring-porous spring xylem growth precedes budburst. Defoliators whose damage follows foliar development can only affect development of later wood. Growth of diffuse-porous and coniferous species responds more drastically, yet differences in NSC storage make them more vulnerable to mortality as stress accumulates. Ring-porous species resist defoliation-related changes in growth and mortality more than diffuse-porous and coniferous species. These findings apply in general to disturbances that cause spring defoliation and should be incorporated into forest vegetation models.