Predisposing trees to die during drought: How physiology and climate history influence mortality in southwestern U.S. piñon pine

Detailed physical and chemical studies of trees that die and survive during drought provide insight into the historic conditions and physiological mechanisms that underpin episodes of tree mortality. We seek to deduce key physiological parameters that influenced the mortality and survival of piñon pine (Pinus edulis) during relatively warmer (2000's) and cooler (1950's) droughts in the southwestern U.S. Using recently sampled and archived tree-ring specimens of trees that died and survived during the 2000's and 1950's droughts, we constructed time series from two measurements that serve as integrated parameters of whole-tree physiological function: radial growth and stable carbon isotope ratios (δ13C) in tree-ring cellulose. We focused our efforts on addressing two hypotheses: 1) piñon trees that died had hydraulic characteristics that predisposed them to succumb to drought and associated bark beetle outbreaks through either hydraulic failure or carbon starvation; 2) dead tree growth and δ13C should be more responsive to climate than surviving trees if drought and temperature driven stress are the ultimate drivers of tree death. We further hypothesize that if dying trees respond to drought by closing their stomata to avoid hydraulic failure, they limit their ability to photosynthesize and should exhibit lower growth and more positive δ13C than trees that survive. Conversely, if trees that died maintained relatively open stomata and succumbed to drought via hydraulic failure, they should exhibit more negative δ13C than trees that survive. Leading up to the 2000's drought, growth in trees that died was lower than in surviving trees, consistent with our expectation. δ13C in trees that died was more negative than in surviving trees but both growth and δ13C of trees that died were less responsive to climate than surviving trees, counter to our initial hypotheses. Our δ13C results are consistent with the hypothesis that the trees that died maintained higher stomatal conductance during drought and hence ran a greater risk of hydraulic failure than trees that survived. An alternative hypothesis is that trees that died had lower leaf area per unit root area, which could also explain higher growth in surviving trees. However, this explanation is inconsistent with previous published results on leaf area of piñon pines that died and survived during the 2000's drought. Notably, tree growth data extending back further in time reveals that growth of live and dead trees was similar until the 1950's drought, when growth trajectories diverged. The mechanism underlying this growth divergence is not yet clear, but this result suggests an important role of previous droughts in predisposing trees to die during subsequent droughts. This raises the possibility of increased future mortality if drought frequency increases even in the absence of changes in drought severity. We are currently expanding our analyses to trees that died and survived across multiple sites covering both the 2000's and the 1950's drought, which will allow us to compare mortality mechanisms across landscapes and between droughts of different magnitudes and temperature profiles.