Sergent, A.S., Varela, S.A., Barigah, T.S., Badel, E., Cochard, H., Dalla-Salda, G., Delzon, S., Fernández, M.E., Guillemot, J., Gyenge, J., Lamarque, L.J., Martinez-Meier, A., Rozenberg, P., Torres-Ruiz, J.M., and Martin-StPaul, N.K.
• Xylem embolism resistance was estimated in 11 species with contrasted wood anatomy and vessel length. • All methods tested provided consistent results in short-vessel angiosperm species. • The Pneumatic method underestimated embolism resistance in conifers and 2 long-vessel angiosperms. • Air-injection and Flow-centrifuge methods consistently estimated embolism resistance intraspecific variability in a conifer. • Turgor loss point measurements were used to evaluate the reliability of vulnerability curves. Vulnerability to drought-induced embolism is a key trait that shapes drought resistance and that could be increasingly used to design climate-smart forest management guidelines and to anticipate the outcome of climate change on populations dynamics and ecosystems functioning. A panel of methods is currently available to measure embolism resistance. This makes crucial a proper identification of which methods are the most accurate for determining this trait. However, the measurement of embolism resistance is sensitive to numerous artifacts that may lead to large errors for a given species. In addition, not all methods are easily accessible because of the cost of some large equipment and/or certain lab facilities. The emergence of the easy and low cost Pneumatic method allows to perform vulnerability curves at high throughput. However, only few studies have evaluated the reliability of this method compared to others. In this study, we proposed a comparison of five methods that allowed to assess embolism resistance in eleven tree species with contrasting xylem anatomy and vessels length (six short vessel angiosperms, two tracheid bearing conifers and three long-vessel angiosperms), covering a large part of the range of embolism resistance observed in trees. Consistent results were obtained among all the methods for short-vessel angiosperm species. In tracheid-bearing conifers, the Pneumatic method overestimated vulnerability to embolism. In long-vessel species, the Pneumatic method led to inconsistent results with accurate vulnerability to cavitation curves (VCs) for one species but led to r-shaped VCs with a underestimation of incipient embolism for the two other ones. The comparison of VC parameters with turgor loss point is proposed as an indicator of the validity of the VCs. The conditions of validity, the advantages and pitfalls of the five methods are discussed. Our results warned against the widespread usages of some methods before rigorous validation tests have been performed. [ABSTRACT FROM AUTHOR]