Patterns of climate sensitivity and memory differ between trees that survive and trees that die during drought

Recent increases in the intensity, frequency, and spatial extent of drought have caused forest dieback in regions around the world. Understanding what predisposes trees to drought-related mortality is becoming a major concern. Leveraging a dataset of ring-width measurements obtained from trees impacted by drought, I attempt to detect early warning signs of drought-related tree mortality in the form of altered climate sensitivity and memory in trees that succumbed to drought. By fitting a Bayesian, mixed-effects model implemented in the stochastic antecedent modeling (SAM) framework to tree growth and climate data, I was able to assess differences in climate sensitivity, as well as the length, strength, and temporal pattern of climate memory in three ecologically relevant tree species: Abies lasiocarpa, Pinus sylvestris, and Populus tremuloides. Differences in climate sensitivity between living trees and trees that later died were species-specific but living trees of all three species showed increased sensitivity to precipitation when temperatures were high, suggesting that they are better able to utilize antecedent precipitation to mitigate the impacts of high temperatures. The length of climate memory did not differ much between living and dead trees, but the species-specific patterns in the importance of the timing of antecedent climate showed trees that survived and those that later died were sensitive to precipitation and temperature at different times of year. These results indicate that clear differences exist in the climate sensitivity and climate memory of trees that survive and those likely to die during drought, which may help identify those trees that are particularly vulnerable