Revised method for constructing acoustic vulnerability curves in trees.

During drought, the formation of air bubbles known as embolisms in the water-conducting xylem reduces hydraulic conductivity, which can ultimately result in tree death. Accurately quantifying vulnerability to embolism formation is therefore essential for understanding tree hydraulics. Acoustic emission (AE) analysis offers a non-destructive method to monitor this process, yet the interpretation of captured signals remains debated. In this study, we introduce an improved methodology for constructing acoustic vulnerability curves (VCAE) that minimizes subjectivity and enhances the accuracy of assessing a tree's vulnerability to drought stress. Our approach combines AE signal clustering with an objective method for pinpointing the endpoint (point of 100% embolism) based on the observed correlation between water potential at maximum AE activity and 50% loss of hydraulic conductivity. By applying a refined clustering algorithm to four temperate tree species (Platanus × acerifolia (Aiton) Willd., Betula pendula Roth, Quercus robur L. and Fagus sylvatica L.), we consistently identified natural frequency-based clusters that effectively separate embolism-related AEs from other signals. This focus on embolism-related AE activity allowed us to minimize the influence of non-embolism-related signals and identify the true VCAE endpoint. Our method, by reducing the subjectivity inherent in previous approaches, enhances the accuracy of VCAE construction, offering broader insights into tree hydraulics and expanding its applicability across different species and environmental conditions.