A major challenge of the coming years will be to maintain biodiversity under changing environmental conditions. Anticipation is an important dimension to tackle this challenge, where models play a major role. So far however, our capacity to predict communities from single species has remained limited. We now need to improve existing models and develop new ones to better account for assembly processes. This is the aim of the recently proposed SESAM framework (Spatially-Explicit Species Assemblage Modeling). Arising from this theoretical formulation, I aim to develop, implement, and test an innovative framework to model species assemblage combining many pre-existing approaches to biodiversity prediction to produce improved spatially explicit projections and overcome single methods limitations. As efforts are currently made on plants, this project will be the first to attempt this comprehensively on animal assemblages, here European reptiles, in 3 nested study areas. To ensure its feasibility, it will be developed in a step-by-step way: i) gathering species and environmental data; ii) performing modeling analyses (species distribution models) to obtain the potential composition of assemblages filtered only for abiotic constrains; iii) defining macroecological constraints on community properties through macroecological modeling; iv) identifying species interactions and deriving ecological assembly rules (EARs) for the considered assemblages, to be used as biotic filter; v) using data and results previously obtained to unify all components; for this, I aim to develop a highly original step to integrate the identified EARs in the novel predictive process. I will test the robustness and scale-dependence of all component and of the whole framework; and finally vi) applying the framework to derive a new generation of climate change projections for reptiles assemblage at all scales. Project results will be relevant for both theoretical science and conservation biology.