Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Doctorat → Suisse → Using stable oxygen and hydrogen isotopes to assess plant water relations in grasslands exposed to drought

ETH Zürich (2013)

Using stable oxygen and hydrogen isotopes to assess plant water relations in grasslands exposed to drought

Prechsl, Ulrich E.

Titre : Using stable oxygen and hydrogen isotopes to assess plant water relations in grasslands exposed to drought

Auteur : Prechsl, Ulrich E.

Université de soutenance : ETH Zürich

Grade : Doctor of Philosophy (PhD) 2013

Understanding the soil-plant-atmosphere continuum as a part of the water cycle becomes an essential task under global climate change. Particularly as the climate variability and the frequency of extreme climate events, such as drought spells, will increase very likely by the end of this century. Recent climate models project a significant decrease of precipitation for Central Europe. For the humid northern part of Switzerland, precipitation is projected to decrease by about 20% during spring/summer until the end of this century. This change in water availability during the vegetation period will have a significant impact on agriculture and ecosystems, as water is a key variable in plant life. On the other hand, plants have developed different strategies to cope with water shortage and it is suggested that plants, such as grasses, do adapt their water uptake by shifting to deeper soil layers. Grasslands are a very abundant agroecosystem across the globe and represent the basis for ruminant husbandry. The productive and agriculturally relevant grasslands of Central Europe depend on sufficient water and typically are dominant in regions with high precipitation, such as northern Switzerland. In respect to climate change, it is therefore important to assess how reduced water availability under summer drought will affect the water uptake of grasslands. Stable water isotopes are a powerful tool to investigate plant water relations, as they allow to trace water without interfering with natural processes. Precipitation and soil water represent the water sources of plants, which are driving the isotopic signal of plant water. Therefore, it is important to understand potential factors influencing the isotopic composition of precipitation and soil water for the interpretation of ecohydrological isotope data. This thesis comprises several experiments to assess water uptake of drought-affected grasslands by means of δ18O and δ2 H in plant and soil water. Central part of this thesis was a three-year precipitation manipulation experiment at three different sites in Switzerland. Summer drought was simulated by means of transparent shelters, which were installed between six to twelve weeks per year to reduce the rain input by about 30%. Aim of this thesis was : a) To test the data quality of a simple rain collector (“ball-in-funnel type collector“ ) for isotope analysis (i) under field conditions and (ii) quantify the alteration of δ18O in dependency of time, relative humidity and sample volume (Chapter 2). b) To investigate how simulated summer drought affects (i) water uptake depth and (ii) belowground biomass distribution of a lowland and an alpine grassland throughout the vegetation period (Chapter 3). c) To investigate if soils can (i) alter the isotopic composition of extractable water and (ii) to reveal the soil properties and (iii) mechanisms responsible for this alteration effect (Chapter 4). We could show that the widely used “ball-in-funnel type rain collector” (BiFC) showed no significant evaporative enrichment under field conditions (Chapter 2). A climate chamber experiment revealed that sample volume had the strongest effect on the oxygen isotopic alteration of the rain samples. Small rain events (≤ 2.5 mm) are more sensitive for the evaporative imprint (up to 1 ‰). However, no significant alteration occurred, if samples were recovered from the BiFC within 5 days. Overall, we could show that the BiFC is an appropriate method to collect rain for isotope analysis. Water uptake of grasslands was studied within a three-year ecosystem manipulation experiment at three sites where we simulated summer drought (Chapter 3). To investigate the water uptake depth of the grasslands, δ18O (δ2 H) values from cryogenically extracted root crown water and soil water from different depth were analyzed by means of two different approaches : (1) linear interpolation method, (2) Bayesian calibrated mixing model. Both approaches revealed independently that the drought-affected lowland grassland used primarily shallow soil layers (0-10 cm) for water uptake throughout the vegetation period, while the plants receiving all precipitation (control) shifted to deeper soil layers (>10 cm). Belowground biomass distribution supported the shallow water uptake. We could not find an increase of belowground biomass in deeper soil layers but rather an increase in the top layer (0-15 cm). The alpine grassland, by contrast, did not show a direct drought response. In an experiment under controlled conditions, we could show that soils can alter the isotopic composition of extracted water (Chapter 4). Oven dried soils, differing in their physicochemical properties, were re-wetted with water of known isotopic composition, which was subsequently extracted by cryogenic water distillation and analyzed for its isotopic composition. Relative to the original water added, extracted soil water was significantly altered by around 1‰ in δ18O. In a similar experiment, where we separated soils in different grain size fractions, we could show that the alteration effect increases with decreasing particle size. Further experiments revealed that soils possess an isotopic memory effect and that matrix potential and equilibrium exchange play a central role in the isotopic alteration of soilwater.



Version intégrale (3,89 Mb)

Page publiée le 29 mars 2015, mise à jour le 20 novembre 2017