MSCA THrESholds

Tropical forests and savannahs are home to a fifth of the world's population, who depend heavily on these ecosystems for their livelihoods. With global change underway, predicting how these ecosystems will react to changes in rainfall and relative humidity is a priority.

Their ecological functioning as alternative stable states means that their response to environmental change is non-linear and particularly difficult to predict. This is due to feedbacks between vegetation structure and fire that stabilise each state, even if environmental conditions change, up to a certain threshold. Once this threshold is crossed, there is a transition from one state to another, and it is virtually impossible to return to the initial state.One of the main limitations to identifying these thresholds is that the ecological processes involved occur over long time scales (>100 years), which exceed the temporal depth of the empirical data available, which is often limited to a few decades.

Fossil archives contained in lake sediments, such as the ones found in the Mediterranean, help to overcome this limitation. However, palaeoecological reconstructions represent a real methodological challenge, because bio-indicators enable us to estimate the variables of interest in the past in a rather qualitative way (more/less), but identifying the thresholds in environmental conditions at the origin of vegetation transitions requires quantitative reconstructions of past vegetation and climatic variables.

The Thresholds project has two main objectives:

• participate in the methodological development of a new quantitative indicator of past relative humidity based on the analysis of the triple isotopic composition of oxygen (17O-excess) in phytoliths, an indicator recently developed at CEREGE
  
  
• to identify the relative humidity thresholds at the origin of forest-savanna transitions, and the ranges of values that have allowed a certain resilience of the forest over the last 5,000 years in Africa, using quantitative reconstitutions of relative humidity and comparisons with reconstitutions of vegetation and fires.

The results of this project will help to better identify and quantify the relative humidity thresholds associated with vegetation transitions. Ultimately, the results of this project will provide a better understanding of the ecological functioning of tropical forests and savannahs, and in particular their response to past environmental changes, in order to better integrate them into vegetation models and provide more accurate projections of their responses to current climate change.

This project was carried out within the framework of and with the technical and material support and collaboration of ANR HUMI-17 and PAST-17.