Borehole equilibration 2.0 or how to chase tracer pulses in a forested ecosystem

Identifying tree water sources has long been an issue since obtaining samples was labor intensive and lacked high time resolution because of the destructive sampling procedure. It was previously shown that the “borehole equilibrium method” (Marshall et al. 2020) allows in situ measurements of xylem sap isotopic composition. While the advantage in using this method is its ability to monitor isotopic composition of xylem continuously and rapidly with immediate data availability, disadvantages are the limited number of trees that can be observed and that the laser has to be present in the field. Here, we propose cheaper and more field-deployable elaboration of the method based on the same principle as to use for tracer pulse-chasing experiments in forested ecosystems.

We installed boreholes in tree stems and sealed them on both sites using brass fittings with a pierceable chlorobutyl septum. The water vapor inside the sealed borehole was assumed to reach isotopic equilibrium with the liquid water in the xylem due to diffusion within seconds and was sampled using gas-tight syringes. The 20ml sample was then injected in a dry air stream connected to a Picarro L2130-I cavity ring-down absorption spectrometer (CRDAS). Standards of known isotopic composition were injected the same way. The peaks, rather than plateaus, of isotopic ratios measured from these injections were weighted by the water vapor amount, giving results accurate enough to distinguish between xylem water of natural abundance and water enriched in deuterium (average SD for ²H 5.2‰ and ¹⁸O 1.9‰ for natural abundance samples). To test this method in the field, we labeled 1m² of soil at different soil depths with 15.5 L of water enriched in ²HHO (δ²H +220000 ‰) in a Scots pine forest in northern Sweden. Trees within a 10m radius from the labeled center were monitored continuously, allowing daily measurements of up to 120 trees for six weeks. Depending on soil depth the uptake dynamics varied over time, with the peaks from the shallowest soil injections  occurring within two weeks, while for the deeper soil layers the contribution to transpiration lagged behind approx. four weeks, likely due to a combination of lower root density and reduced hydraulic conductivity at greater depth.  The strength of the peaks was correlated with distance from the labeled soil patch.

We were able to show that this method works to chase an artificially enriched water pulse through a natural forested ecosystem. At the same time, this adaptation allows the method to become even cheaper than its precursor as it requires much less tubing and fewer fittings. Lastly, we consider it more field-deployable because it does not require the CRDAS to be in the field.