Thesis project : reliability of species distribution model projections

Climate change is causing profound changes in the functioning of ecosystems and the geographical distribution of species. Natural environment managers urgently need reliable projections of species range changes across their territories and beyond. As climate change is currently in a phase of acceleration, it is urgent to determine in a clear and precise way the robustness of the models that are intended to provide such projections, and to identify the origins of their robustness.

This thesis project is fundamentally aimed at improving the credibility that managers and decision-makers place on the projections generated by species distribution models for the coming decades. To achieve this goal, we propose a fine comparison of the capacity of correlative models and process-based models to predict past and future distributions of iconic temperate tree species from European forests (Abies, Betula, Corylus, Fagus, Fraxinus, Larix, Picea, Pinus, Quercus). This comparison will involve three temporal situations: distant past (-12000 years), recent past (-50 to 100 years), near future (2025-2100) and different versions of the models that will differ by their level of complexity and the methods of estimation of their parameters in order to test different hypotheses on the origin of their robustness or lack of robustness.

Thesis supervisor : Isabelle Chuine - Research team : Forecast

Starting date : October 1st, 2021

Funding : PhD fellowship, GAIA Doctoral School