Daniella M Rempe, Erica L McCormick , W Jesse Hahm, Geeta G Persad, Cameron Cummins, Dana Ariel Lapides , K. Dana Chadwick, David Nicholas Dralle 

Root-zone water storage (RWS) dynamics regulate when plants experience drought-related water stress and mortality. However, because RWS capacity (Smax) is poorly known, it remains challenging to translate variability in precipitation to water stress. Here, we investigate the relationship between precipitation variability and Smax and implement a framework for identifying the vulnerability of seasonally dry woody ecosystems to projected climate change. Using novel estimates of Smax across California, we demonstrate that where dry-season RWS is routinely capped by Smax, plants are less vulnerable to precipitation variability relative to where dry-season RWS varies annually with precipitation. Using direct measurements of RWS and Smax at three field sites, we illustrate how these differences in vulnerability arise due to variations in bedrock properties. We calculate that up to 23% of California's total biomass is sensitive to year-to-year variations in precipitation and can experience carryover of moisture from one year to the next. Contrary to the notion that deep weathering and moisture carryover confer ecosystem resilience to moisture stress, the areas we identified where Smax commonly exceeds precipitation totals experienced disproportionately high rates of mortality during recent drought. In contrast, the 51-58% of California's total biomass that experiences annually reliable dry-season moisture supply showed lower drought-related mortality. This framework then allows us to use climate projections for the next century to determine that a transition from stable to unstable moisture supply is projected for 3% of the state's carbon stocks. Much of the area presently showing signs of vulnerability is expected to experience additional moisture stress in the coming century due to changes in precipitation amounts alone. An understanding of belowground conditions, including the deep root-zone in bedrock, contributes to prediction of conditions leading to ecosystem water stress.