Tree water deficit as a drought stress indicator

The frequency and severity of droughts is expected to increase as climate change develops, which negatively affects forests and their numerous benefits. However, some species may be adapted better to the new conditions than others. Hence, evaluating the stress level of trees under ranges of water-limited conditions is crucial to judge and predict forest health. While water potential measurements are valuable for indicating stress, they are hardly applicable to monitoring continuous developments or investigating large numbers of individuals simultaneously. An alternative attempt is to utilize high-resolution dendrometer measurements of stem shrinkage, which is caused by the reduction of the tree's internal water storage (tree water deficit; TWD) and is thus an indirect indicator of stress. There is first proof of correlations between TWD and water potential under moderate drought stress. However, we still lack observations that relate stem diameter variations to severe drought stress when embolism formation (air bubbles in the xylem preventing water flow) occurs. Also, recovery responses after re-wetting, and the respective linkages to physiological states and processes are fairly unknown.Here, we present results from a greenhouse experiment on potted tree saplings. Two widespread temperate conifers (Pinus sylvestris,Larix decidua)were exposed either to drought-recovery cycles or to lethal drought until complete dehydration under controlled experimental conditions. We found strong relations between TWD and both midday water potential and gas fluxes across the full range of dehydration, with only minor differences between the two species. Re-wetting after a short drought period had no effect. Conversely, re-wetting after a drought severe enough to cause hydraulic damage significantly affected the correlations due to different recovery times.Our results indicate the great potential of dendrometers to provide continuous and cost-efficient time series that allow valuable insights into the water status and thus drought stress of trees. While all stages of dehydration can be covered, the dependencies of re-wetting responses after severe droughts are still unclear and more species-specific investigations are required. Nevertheless, applying TWD seems to be a promising way forward to improve our understanding of drought-stress-induced forest decline and drought-recovery dynamics.