The study, led by Youri Rothfuss of Forschungszentrum Jülich together with colleagues in Germany, Belgium and at the University of Bonn, presents a proof of concept for continuously monitoring the stable isotope composition of water inside the stem of a living, herbaceous plant — something previously achieved only for trees. The team drilled a small hole, 3.5 millimeters wide and 5 millimeters deep, into the stem of sunflower plants a couple of centimeters above the soil surface, sealed a thin PTFE sampling tube into the borehole, and connected it to a cavity ring-down laser spectrometer. Water vapor diffusing from the stem tissues was drawn through the tube at 28 milliliters per minute and analyzed for its hydrogen and oxygen isotope ratios every single second, giving researchers a real-time window into the plant’s internal plumbing.

Before any field-relevant conclusions could be drawn, the method had to be calibrated. In a first experiment, nine sunflower plants of the compact “Yellow Spray” cultivar were grown hydroponically for three months and then transferred into water baths with three distinctly different isotope compositions: one strongly depleted, one matching local tap water, and one strongly enriched. Once the stem vapor readings reached a stable plateau — which took between 20 and 45 minutes — the researchers could establish linear calibration functions linking the vapor measured in situ to the liquid water actually circulating in the xylem. The fits were remarkably tight: more than 99.9 percent of the variability in the hydroponic solutions was explained, with root mean square errors of just 1.8 permil for hydrogen and 0.8 permil for oxygen. Notably, in a separate soil-grown test, the response time of the isotope signal dropped to a mere 20 to 30 seconds — essentially the response time of the laser spectrometer itself.

With the calibration in hand, the team moved to the centerpiece of the study: two sunflowers grown individually in soil columns inside a climate-controlled chamber, where temperature, humidity and light were held under strict control. To independently verify what the isotope probe was seeing, the researchers deployed the Soil Water Profiler, or SWaP, a capacitive sensor developed at Jülich that slides up and down along the soil column on a precision axis, measuring volumetric water content at centimeter resolution with an accuracy of up to 0.003 cubic centimeters of water per cubic centimeter of soil. Because changes in light intensity rapidly alter transpiration and root water uptake while vertical water redistribution in the soil proceeds more slowly, the local rate of water content decline can be read as a direct measure of how much water the roots are extracting from each depth.

The experimental design was elegant in its simplicity. The researchers injected pulses of isotopically labeled water into specific depths of the soil — enriched water near the surface, depleted water at 30 centimeters depth through a porous aeration stone — and then simply watched. When enriched water was added to the top layer of the first plant’s column, the contribution of the upper soil layer to total root water uptake surged while uptake from the middle layer dropped, and within minutes the isotope composition of the stem water began to rise. When depleted water was later injected deep into the profile, the pattern reversed: uptake from the middle layer jumped from 0.06 to 0.95 milliliters per 10 minutes while surface uptake collapsed, and the stem isotope signal turned over in lockstep. The two independent techniques — isotope monitoring and soil water profiling — told the same story, validating the new approach against an established one.

But the most striking finding emerged when the measured isotope dynamics were compared with steady-state model calculations. A classical mixing model, which assumes that water extracted from the soil instantly and completely mixes inside the roots and stem, predicted instantaneous jumps in stem water isotope composition following each change in uptake pattern. The real plants did nothing of the sort. Even under high transpiration demand, the observed isotope values took roughly two hours to approach a new plateau, and after a full night of uptake the stem water had still not fully equilibrated with the labeled soil water. Something inside the plant was slowing the signal down — and the culprit was the stem itself.

Using a simple first-order replacement model, the researchers estimated that only about one sixth of the water flowing through the xylem exchanges with the water stored in the stem’s non-conducting tissues, the pith and cortex. With a total root water uptake of roughly 3 milliliters per hour and a sampled stem volume of about 2 milliliters, this exchange fraction implies that it takes around ten hours for the stem reservoir to be nearly fully replaced by newly absorbed water. In other words, when the sun opens its stomata and transpiration spikes faster than the roots can respond, the sunflower does not immediately draw the difference from the soil — it mobilizes water stored in its own stem, much like a camel drawing on its hump. When transpiration falls, the surplus is redistributed back into those tissues. The stem, in effect, functions as a rapidly accessible buffer, a behavior long recognized in trees but rarely quantified in herbaceous crops.

This finding carries real consequences for the wider scientific community. Isotope-based studies of plant water sources — spanning agriculture, ecology and hydrology — routinely treat stem water as a faithful snapshot of current root uptake. The new results show that this assumption breaks down under dynamic conditions, particularly when light intensity changes or when the soil dries out. The isotopic composition of water in aboveground tissues, the authors argue, should not be read as a static map of belowground reservoirs but as a dynamic signal shaped by the plant’s ever-changing water demand. For large woody species, whose stem reservoirs are enormous relative to their sap flow, the distortion could be even more pronounced — a caveat that recent work on tree sap and tissue water has already begun to flag.

The method itself is strikingly accessible. Beyond the laser spectrometer, it requires no exotic equipment: a drill, a thin tube, and sealing paste. The plants showed no visible signs of stress during the experiments, and magnetic resonance imaging of the stems after the trial confirmed the drilling left only localized damage. Independent pre-tests suggest continuous monitoring is feasible for at least 96 hours, although the authors caution that long-term applications — and possible plant defense responses that might seal the borehole — remain to be tested. The technique proved sensitive to the exact insertion depth and height of the tubing, working best when the tube penetrated only a few millimeters into the stem close to the root crown, and calibration functions will need to be re-established for different growth stages and likely for other species.

The researchers are explicit about the limits of their proof of concept: with no experimental replicates, the study was not designed to draw definitive conclusions about sunflower physiology in general. Yet the combination of continuous stem isotope monitoring and non-destructive soil water profiling opens a door that ecohydrologists have been pushing at for years. It promises a way to quantify plant hydraulic capacitance directly, to study the residence and transit times of water inside plants in the same way isotopes have long been used to trace water through watersheds, and to reconcile the isotope compositions of root uptake, xylem water and transpiration within a single coherent framework. As droughts intensify across the world’s croplands, understanding how plants bank water in their own bodies — and how quickly they can draw on that bank — may prove to be far more than an academic curiosity.

Subject of Research: In situ xylem water isotope monitoring and root water uptake dynamics in sunflower

Article Title: Online xylem water isotope monitoring and soil water content profiling reveal spatial root water uptake dynamics in sunflower

Article References: Rothfuss, Y., Le Gall, S., Brüggemann, N., Jahan, S., Javaux, M., Klaus, J., Vereecken, H., & van Dusschoten, D. (2026). Online xylem water isotope monitoring and soil water content profiling reveal spatial root water uptake dynamics in sunflower. Hydrology and Earth System Sciences, 30(18), 6095-6114. https://doi.org/10.5194/hess-30-6095-2026

Image Credits: AI Generated

DOI: 10.5194/hess-30-6095-2026

Keywords: sunflower, xylem water, stable isotopes, root water uptake, plant hydraulics, stem capacitance, laser spectrometer, soil water content, Soil Water Profiler, drought stress, ecohydrology, Hydrology and Earth System Sciences

The research, published in the journal Plant and Soil by Shogo Imada and Masaru Nagai of the Institute for Environmental Sciences in Rokkasho and their colleague Hideki Kakiuchi of the institute’s Tritium Research Center, set out to test a deceptively simple idea. If natural tritium concentrations in soil water show a clear vertical pattern near the surface, then the tritium signature of water inside tree branches could reveal which soil layers the trees have been tapping. Between the summers and autumns of 2023 and 2024, the team collected soil cores and branch samples from overstory trees at study sites in northeastern Japan, working with three species that dominate humid temperate forests of the region: the deciduous oak Quercus crispula and two pines, Pinus densiflora and Pinus thunbergii.

Tritium is the radioactive heavy isotope of hydrogen, and in the environment it exists overwhelmingly in the form of tritiated water, chemically indistinguishable from ordinary H2O. It is produced naturally in the upper atmosphere by cosmic ray interactions and has also been released by atmospheric nuclear weapons testing in the mid-twentieth century and by nuclear facilities. Because tritium is part of the water molecule itself, it moves with water through soil and plants without the complications that affect other tracers. Unlike the stable isotopes deuterium and oxygen-18, tritium does not undergo significant fractionation during most biological processes, meaning its concentration in xylem water should faithfully reflect the mixture of soil water sources a tree has absorbed. That property makes it an attractive candidate for tracing water movement through the soil-plant-atmosphere continuum.

The technical challenge is that natural tritium concentrations in today’s environment are extremely low, decades after the peak of bomb testing. To measure them, the researchers extracted water from soil and branch samples using cryogenic vacuum extraction, a technique in which samples are frozen and then heated under vacuum so that the water sublimates and condenses in a cold trap. Previous work has shown that this extraction method barely alters low-level tritium concentrations in the free water of environmental samples, which was a critical prerequisite for the study. The extracted water was then subjected to electrolytic enrichment, a process that concentrates tritiated water molecules relative to ordinary water by preferentially electrolyzing the lighter isotopologues, before the enriched samples were counted with liquid scintillation techniques to determine tritium activity.

The results were striking. At both study sites, the team observed clear vertical gradients in soil water tritium concentrations within the uppermost soil layers, within the top 10 centimeters at one site and the top 20 centimeters at the other. The gradients were not random. Tritium concentrations generally increased toward the soil surface as soil moisture decreased within these depth ranges, a pattern that points to evaporative enrichment as the likely mechanism. When water evaporates from the soil surface, remaining water becomes progressively enriched in heavier water molecules, including tritiated water, much as a drying puddle becomes saltier at its edges. The researchers suggest that this evaporation-driven process shapes a tritium profile steep enough to serve as a usable depth marker even in the isotopically homogenized soils of humid temperate climates.

With a usable soil profile in hand, the team turned to the trees. Tritium concentrations measured in the xylem water of the oak and pine branches closely matched those of soil water at depths shallower than 25 centimeters. That correspondence implies that the overstory trees at these sites were drawing the bulk of their water from the near-surface soil layers during the study periods. The finding is notable because it runs counter to the common assumption that large forest trees necessarily rely on deep water sources. Other studies in temperate hardwood forests have similarly found substantial reliance on shallow soil water, and the new tritium evidence adds an independent line of support in an environment where the conventional stable isotope approach struggles.

The broader context makes the result more than a methodological curiosity. Understanding the water uptake depth of forest trees is central to predicting how forests will respond to drought, how species coexist by partitioning belowground resources, and how water moves from soil through vegetation back to the atmosphere. Transpiration by plants is a major flux in the global water cycle, and models of land-atmosphere exchange depend on assumptions about rooting depth and water source that are difficult to verify. Global syntheses of tree water uptake patterns have revealed enormous variation across biomes, but humid temperate forests remain underrepresented, precisely because the standard tools fail there. A tracer that works where stable isotopes do not could fill a significant gap in the map of global vegetation hydrology.

Tritium has, of course, a long history in hydrology, where its radioactive half-life of about 12.3 years made it invaluable for dating young groundwater and tracing recharge in the era of bomb-era tritium fallout. Its application to plant water sourcing has been far rarer, partly because of the analytical difficulty of measuring the very low concentrations found in today’s precipitation and soils. The Japanese team’s demonstration that electrolytic enrichment combined with cryogenic extraction can resolve meaningful gradients in soil water and match them to xylem signatures suggests that the technique could be adopted more widely, particularly in regions with monitoring infrastructure for environmental tritium. The authors note that distinct vertical gradients in natural tritium can occur in near-surface soil water in humid temperate forests and may provide useful information for inferring tree water uptake depths, a conclusion that opens the door to comparative studies across sites and seasons.

Cautions remain. The study covered two growing seasons at sites in northeastern Japan, and the observed gradients depended on evaporative enrichment, which may weaken under prolonged wet weather or differ across soil textures. The researchers also note that interactions between tritiated water and soil minerals, such as adsorption and exchange in clay-rich horizons, could complicate profiles in some settings, although such effects appear limited in the near-surface layers examined here. Seasonal variation in precipitation tritium inputs adds another layer of variability that future work will need to quantify. Still, the study demonstrates that a tracer once valued mainly for tracking fallout and groundwater age can be repurposed to answer one of ecology’s most persistent hidden questions: how deep the roots of a forest actually reach. As climate change intensifies drought stress in even the wettest temperate forests, knowing where trees drink may prove essential to forecasting which forests endure and which fade.

Subject of Research: Use of natural tritium as a tracer to estimate the water uptake depth of trees in humid temperate forests

Article Title: Effectiveness of the use of natural tritium as a tracer to estimate the water uptake depth of humid temperate trees

Article References: Imada, S., Nagai, M., & Kakiuchi, H. (2026). Effectiveness of the use of natural tritium as a tracer to estimate the water uptake depth of humid temperate trees. Plant and Soil. https://doi.org/10.1007/s11104-026-09089-z

Image Credits: AI Generated

DOI: 10.1007/s11104-026-09089-z

Keywords: tritium, water uptake depth, humid temperate forest, soil water isotopes, xylem water, cryogenic vacuum extraction, electrolytic enrichment, evaporative enrichment, root water uptake, forest ecology, Japan, Plant and Soil