Climate models predict numerous changes in tropical forest regions, e.g. increasing frequency of drought events (Duffy et al. 2015) and increasing fertilization by atmospheric CO2 concentration modulated by soil nutrient limitations (Terrer et al. 2019). These changes could lead to deep disturbances of the Amazonian forest (Huntingford et al. 2004). Even if these projections remain uncertain, mature tropical forests have been found to be very vulnerable to the observed increase in extreme drought events over the past decades (Phillips et al. 2009). Recently, we demonstrated that climate change induced a shift in community composition in permanent forest plots across Amazonia (Esquivel-Muelbert et al. 2019) and found that a transition towards a more drought-tolerant Amazonia is under way. In this line, establishing tree species response distribution in regards to climate change appears a necessity in Amazonia.

At the community level, morpho-anatomical functional traits have been largely used to explain environmental filtering (Engelbrecht et al. 2007; Kraft et al. 2008; Fortunel et al 2014). However, these provide limited mechanistic understanding of plant response to environmental constraints and distribution (Paine et al. 2018). In order to identify changes to species distribution within climatic niches, intraspecific trait variability (ITV) could be used to explain response to abiotic gradients (Violle et al. 2012) and habitat association rather than species mean trait values only. Habitat association is not necessarily identified by the species geographic range size, nor by climatic niche breadth, or by intraspecific variability of morpho- anatomical leaf traits (Fortunel et al. 2019). These results emphasize the need to look for more appropriate physiological key traits, more closely linked to environmental constraints (climate and soil). Mechanistic traits that are closely linked to species physiological responses to abiotic gradients (Delzon 2015, Brodribb 2017) offer a promising way forward to better explain current tree species distribution and predict their change due to climate change.

This project aims to predict changes in species distribution to climate change using species’ hydrological affinity and mechanistic traits. We will achieve this by combining: current species distribution, hydrological indices, climatic predictions, and use a few keys traits at the individual level in innovative models (joint species distribution model; JSDMs). We will use innovative models to better understand the processes by which species interact with their environment. Finally, we will create a map of future species distribution under climate change.