Tracing tree water uptake with stable isotopes depends on accurately characterizing the isotopic composition of plant source water in both oxygen and hydrogen. Cryogenic vacuum distillation (CVD) is the standard extraction method, but recurring discrepancies between CVD-extracted water and expected source water values have raised questions about its reliability, particularly for hydrogen.

In our new study, David N. Steger, Eligio Amicabile, Daniel B. Nelson and Ansgar Kahmen ran two irrigation experiments on juvenile temperate trees, comparing CVD-extracted soil, xylem, and phloem water against irrigation water and mildly vacuum-extracted sap as an isotopically accurate reference. While our results hint at the existence of two isotopically distinct soil water pools, further physiological processes did not substantially alter our findings, and overall method validity was not compromised. CVD proved a reliable extraction method for δ18O-based source water identification, whereas δ2H values carry systematic artefacts that require correction before use in ecohydrological studies.