Thresholds in past relative Humidity and Ecosystem Structure in Africa

Tropical forest and savanna are home to one-fifth of global human population.

With ongoing global change, predicting how these biomes will respond to changing rainfall and relative humidity (RH) is of high priority.

The overall aim of this project is to identify thresholds in RH that triggered in the past (last 10,000-years) the transition from forest to savanna and those that made forest resilience possible in Africa.

I will use a new RH proxy, recently developed by the host laboratory, which is the triple oxygen isotope composition (17O-excess) of phytoliths.

While calibrations were conducted in growth chambers, field applications are needed to precisely evaluate the accuracy of the proxy.

I will examine the evolution of the 17O-excess in the water cycle and in grass and tree phytoliths, and quantify the isotopic fractionation at play, at a West African instrumented site.

Then 17O-excess bias potentially brought by the contribution of phytoliths from lake shoreline vegetation to bulk sedimentary phytoliths will be quantified.

Finally, I will analyse the 17O-excess of phytoliths from four West and Central African Holocene lacustrine sedimentary cores to reconstruct changes in RH. Vegetation structure and fire activity will be reconstructed from the same sediments.

The effect of reconstructed RH and fire activity trends will be analyzed using time series cross-correlation and spectral analyses, and thresholds in RH identified using regime shifts analyses. These results will help to better understand the impacts of past climate change on tropical forest resilience.

During this project, I will acquire new laboratory skills, develop the use of a new paleoclimate proxy, reinforce my scientific leadership and expand my career prospects in academia. I will bring my expertise in remote sensing analyses and biome management to the host laboratory.

This project will also strengthen the European leadership in triple oxygen isotope analyses for paleoclimate reconstructions.