The strength of the terrestrial carbon sink—dominated by forests—remains one of the greatest uncertainties in climate change modelling. How forests will respond to increased variability in temperature and precipitation is poorly understood, and experimental study to better inform global vegetation models in this area is urgently needed. Necessary for achieving this goal is an understanding of how increased temperatures and drought will affect landscape level distributions of tree species. Quantifying physiological thresholds representing a point of no return from drought stress, including thresholds in hydraulic function, is critical to this end. Recent theoretical, observational, and modelling research has converged upon a threshold of 60 percent loss of hydraulic conductivity (PLC) for mortality. However, direct experimental determination of lethal points in conductivity and cavitation during drought is lacking. We quantified thresholds of hydraulic function in loblolly pine, Pinus taeda, an ecologically and commercially important species. In a greenhouse experiment, we exposed saplings (n=83) to drought and then re-watered them at levels of increasing water stress determined by pre-selected targets in pre-dawn water potential. We measured physiological responses, including hydraulic conductivity, native PLC, foliar color, and dark-adapted chlorophyll fluorescence. We observed saplings for two months after re-watering to determine survival.
Results/Conclusions
In this study, we directly experimentally tested for the point of no return from hydraulic failure with a drought experiment in loblolly pine (Pinus taeda). By directly measuring PLC across a water-stress gradient ranging from mild to lethal, we observed the lowest lethal PLC to be 70%, and the majority of trees which died (n=47) had greater than 90 PLC. Some trees (n=8) were able to survive hydraulic stress above 90 PLC. We found that the PLC at which half of trees died and half survived was 80%. This threshold for hydraulic failure was much higher than the anticipated 60% suggested by previous theoretical, observational, and inferential studies. At the time of rewatering, foliar color change from pre-drought measurements was predictive of whether trees would survive or die. Our observation raises the question—do other species exhibit similarly high thresholds for hydraulic failure? Additional empirical study of multiple species is needed to understand variability in lethal hydraulic failure thresholds for models presently in development.