Drought influences tree physiology on multiple timescales, from seasonal responses to moisture stress to multi-year legacy effects of drought. Mechanisms underlying drought legacies are numerous, but dynamics in non-structural carbohydrates (NSCs, soluble sugars and starches) likely play a major role. We investigated how water stress at different time-scales—instantaneous (pre-dawn water potential [PDWP]), days to weeks (leaf δ13C), and growing season to years (current year growth)—may influence observed NSC in leaves and stems of trees across sites with varying drought histories. We conducted a field study that spanned 22 sites across the Four-Corners Region of the Southwestern U.S. (AZ, NM, UT, and CO) with pre- and post-monsoon NSC sampling and tree ring collection (>700 cores) of pinyon pine (Pinus edulis), Utah juniper (Juniperus osteosperma), and aspen (Populus tremuloides). We focused on the following questions: (1) How does NSC in different species respond to water stress at different time-scales? and (2) Are inferred drought legacies related to NSC levels in certain species or tissues? In order to evaluate relationships between NSC pools and drought legacies, we modeled tree-ring width as a function of antecedent climate and calculated drought legacies following the regional 2012 drought.
Results/Conclusions
Our random-intercepts model with unique effects of moisture stress indices for season, tissue, and species indicates that NSC levels were most strongly related to moisture stress in aspen across timescales (R2= 0.60), and least in pinyon (R2= 0.20). In aspen, soluble sugars and starch content were related to all indices of moisture stress (p<0.05), but the importance of each index depended on time within the growing season (pre- or post-monsoon) and tissue (leaf or stem). For example, leaf sugars were related to PDWP before monsoon arrival, while stem soluble sugars were related to current year ring width only after monsoon arrival. Drought legacies inferred from the tree-ring model were most strongly related to NSC levels in aspen but only weakly in pinyon and not at all in juniper. Aspen also had the longest climatic memory (only 50% of precipitation response explained by most recent 26 months of climate), suggesting NSCs link past climate and current growth in this species. We suggest NSC variability partially drives legacy effects of drought in species with high NSC storage (aspen), and hypothesize hydraulic damage could be more important in species with lower NSC storage (pinyon and juniper).