The isotope composition of cryogenically extracted stem water of trees does not necessarily equal the isotope composition of the trees source water

The oxygen and hydrogen isotope composition of water in the stems of trees can be used to estimate the origin of source water of trees in the soil-plant atmosphere continuum and it serves as starting point for the interpretation of the oxygen and hydrogen isotope composition of plant organic compounds. These applications rest on the belief that the water extracted from tree stems is identical in its isotopic composition to the trees source water. Recently, there have been indications that this assumption might not be true and that different isotope effects cause unexpected deviations in stem water oxygen and hydrogen isotopes from the trees source water. To identify the mechanisms that cause these effects, we performed a series of greenhouse experiments. We tested (i) if fractionation occurs during the water uptake by roots, (ii) if oxygen and hydrogen isotopes in stem water show the same positive or negative deviations from soil water, and (iii) if isotope effects in different water pools in stem (water in conducting cells and water in living parenchyma cells) can explain deviations in the isotope composition of stem water from source water. To address these questions we grew 1 - 2 m tall saplings from seven different temperate tree species in a greenhouse under stable environmental conditions. We found that stem water indeed deviated from the source water in the soil. Interestingly the average direction of this deviation was different for oxygen isotopes (+0.5 per mil) and hydrogen isotopes (-10 per mil). The analysis of separate water pools in the stems revealed that observed effects are driven (1) by slight positive isotope effects during plant water uptake for both, oxygen and hydrogen. In addition (2) water extracted from parenchyma cells in the tree stems showed for hydrogen a up to 20 per mil depletion in ²H compared to water in conducting tissue (xylem) but no such effect for oxygen. The sum of these effects resulted in the bulk stem water that was slightly ¹⁸O enriched and substantially ²H depleted compared to the trees source water. Our study shows that stem water does not necessarily equal the isotope composition of source water, that these deviations differ for oxygen and hydrogen isotopes, and that these deviations need to be considered in the interpretation of stem water isotope data of trees.