The 17O-excess of plant silica, a proxy of continental atmospheric relative humidity in tropical areas: field calibration

Combined with atmospheric temperature, continental atmospheric relative humidity (RH) can be used to estimate vapor pressure deficit, a driver of soil water evaporation, plant transpiration and ecosystem productivity. This makes continental RH a key-parameter to consider when assessing interactions between vegetation and atmosphere in a changing climate. Accumulative evidence indicates that, at the global scale, continental RH is decreasing since the end of the 20th century and will continue to decrease in relation to global warming. However, simulations of continental RH for the future are highly variable, especially for the inter-tropical area. Model-data comparisons, applicable beyond the instrumental period would help to improve the accuracy of RH estimates, along with our understanding of vegetation-climate interactions.

Within the HUMI-17 project, a new quantitative proxy for RH was recently developed, namely the triple oxygen isotopic composition of plant silica (i.e. phytoliths), expressed by the 17O-excess. Ten per meg of change in phytolith 17O-excess reflects a change in RH of ca. 2 %. The proxy calibrations were derived from growth chamber experiments where the climate parameters varied individually, or from mono-specific crop monitoring in the field. Here,we examined the evolution of the triple oxygen isotope composition of rainwater, soil water, plant water and phytoliths in relation with seasonal changes in climate at two instrumented sites in West Africa (Senegal, Benin), which are part of the AMMA-CATCH eco-hydrological observatory. Climate data were processed at different time scales. The triple oxygen isotope compositions of water and silica were measured by cavity ring-down spectrometry and isotope ratio mass spectrometry. The obtained dataset allows to answer the following questions: Can the phytolith 17O-excess vs RH relationship be used whatever is the vegetation type? Which RH (annual, seasonal, under or above the canopy) is recorded by the 17O-excess of phytoliths from tropical forests and savannas? This calibration study is a necessary step prior to accurate RH reconstructions using the 17O-excess of fossil phytoliths and model-data comparisons. It also enhances our understanding of the evolution of the 17O-excess within the water cycle at the soil-plant-atmosphere interface.