OASIS combined methodological developments with mesocosm and field experiments to investigate soil–plant–atmosphere exchanges in arid and semi-arid ecosystems. A key achievement was the development and deployment of a multi open dynamic chamber system allowing continuous measurements of CO2 and water vapour fluxes under very low soil moisture conditions, where conventional approaches often face strong limitations.
Experiments were conducted on both bare soils and vegetated systems representative of dryland environments, with particular emphasis on drought-adapted vegetation, including plants with a Crassualcean Acid Metabolism (CAM) photosynthetic pathway that absorbs CO2 during nightime. Measurements were designed to capture diel dynamics and to provide a quantitative partitioning of net fluxes into their main components. For CO2, this included distinguishing photosynthetic uptake from respiratory losses, as well as separating autotrophic and heterotrophic respiration. For water vapour, exchanges were analysed by differentiating evaporation, plant transpiration, and atmospheric water vapour adsorption.
A specific focus was placed on nocturnal fluxes, which are particularly relevant for CAM-dominated ecosystems. Current in situ approaches, such as eddy covariance, are known to face limitations at night due to low atmospheric turbulence and reduced sensitivity, particularly under dry conditions where fluxes are often of low magnitude and occur as short-lived pulses that are challenging to capture. The mesocosm-based “dryland ecosystem analog” developed in OASIS provided a controlled framework to capture these nocturnal exchanges and to explore flux dynamics that remain difficult to resolve in the field.
The experimental framework also included the simulation of climate change scenarios, notably heat waves and extreme rainfall events. These manipulations revealed strong and rapid responses of both carbon and water fluxes, highlighting the sensitivity of dryland ecosystems to climatic extremes.
Within the experimental configuration of the project, biological CO2 fluxes associated with vegetation activity and soil respiration dominated observed signals. Geochemical fluxes, such as those potentially related to mineral carbonation, were comparatively small in magnitude and could not be resolved relative to biological variability and instrumental uncertainty.