In-situ measurement of the stable isotopes of soil and xylem water using liquid-vapor equilibration: protection against water intrusion and maximum tubing lengths for automatic systems

The climate-change-induced increased frequency of droughts and shifts in rainfall patterns will most likely impact the interaction of trees with components of the hydrological cycle, e.g., rainfall, soil water, or groundwater. To study this, an increasing number of scientists use in-situ measurement systems capable of high-resolution measurement of the stable water isotopes (δ¹⁸O, δ²H) of xylem and soil water. These systems often use gas-permeable probes to sample water vapor in isotopic equilibrium with the liquid xylem or soil water, which are connected to transport tubing of several meters length that guide the vapor sample to the gas-inlet of a field-deployed isotope analyzer. Potential issues of these systems include (a) the accidental transport of liquid water to the isotope analyzer (e.g., by damage to the tubing, or inadequate sealing of connections), and (b) the maximum transport tubing length to obtain a reliable measurement. Here, we tested two different syringe filters (0.45 µm Nylon, and 0.2 µm PTFE) in terms of preventing liquid water from passing through, and from allowing water vapor to pass through without fractionation of isotope ratios. By switching between two known water sources, we further analyzed the effect of a possible filter cake made up of water vapor of the previous measurement trapped in the filter material on subsequent isotope measurements (memory effect). Lastly, using a 4 mm diameter tube we tested lengths from 1.3 m to 15.3 m in 1-m-increments to assess maximum tubing lengths for a reliable analysis. Results showed that only 0.2 µm filters were able to prevent liquid water from breaking through, and that isotope values were slightly enriched (δ¹⁸O: +0.47‰, δ²H: +1.3‰). However, this enrichment was not statistically significant due to the small sample size of only three repeated measurements with and without the filter installed. No influence of a possible filter cake was found as two waters of known isotope ratios could be repeatedly measured when switching back-and-forth between water sources (standard deviations were on average 0.15‰ for δ¹⁸O and 0.6‰ for δ²H). Tests of tubing length showed a maximum length of 6.3 m for the isotope ratios to reach the target value when measuring for 20 minutes. Between 15.3 m to 12.3 m, no discernible change in isotope ratios was detected, and from 12.3 m to 7.3 m the expected isotope ratio was only detected after the 20-minute measurement window. Using the vapor volume of our 4 mm diameter and 6.3 m long tube of approximately 80 cm³, we calculated that the often-used tubes of 1/8-inch inner diameter (~1.58 mm) could theoretically be up to 40 m long. We thus recommend using a maximum transport tubing length that corresponds to approximately 80 cm³ of gas that needs to be transported. If liquid water intrusion might pose a danger to field-deployed measurement equipment, 0.2 µm PTFE syringe filters can be used to stop the liquid water. However, the issue of potential fractionation of these filters is not yet resolved.