Abstract : Grasslands ensure a variety of services to the agricultural community and the society as a whole. Dairy and beef production are the major objectives of grassland management. However, they are also effective to mitigate greenhouse gas (GHG) emissions by sequestering the atmospheric carbon into the soil. Considering the challenges of demographic and climate changes (CC), anticipating the risks associated to the continuity of these services is to date a major societal concern (and is likely to remain a topic of public concern in the foreseeable future). In this context, the thesis aimed at: i) modelling the impacts of CC on grassland services (including the GHG budget), while accounting for the increased variability of future climate as projected by climate models ; ii) identifying pathways to CC adaptation of pasturebased livestock systems. By considering such goals, novel animal and plant modules were developed for the Pasture Simulation model (PaSim) that allowed, on the one hand, a mechanistic representation of the animal performance and methane emissions at pasture and, on the other hand, assessing interactions between vegetation diversity, CC and pasture management. PaSim has been used to simulate the impacts of CC on monocultures and multispecies grasslands run at 12 French sites across a range of soil, climate and management conditions. The model has been forced by high-resolution climate projections over 1950-2100, as generated by a set of socio-economic scenarios, climate models, and regionalization / initialization methods. Overall, the projected consequences of CC on French grasslands are not severe. Major concerns refer to summer forage deficit and water table depletion, for which significant risks of reduction may arise. If the most arid sites will not necessarily experience the worst CC impacts, they remain affected by the most critical conditions of forage and water deficit. Some opportunities for forage production that may arise for other seasons than summer involve important seasonal changes, which may also require the adaptation of pasture-based livestock production systems. A case study on two major production districts - Brittany and Massif Central - showed that CC will affect accompanied by the need of modifying forage and concentrate distribution to complement the grazed pastures. Moreover summer grazing will not be possible in some years. Flexible systems will allow better exploiting grazed grasslands. This may be achieved by extending the grazing periods and increasing the area allocated to pasture, while also increasing the animal density. Such adaptation measures proved not to worsen the GHG budget of the system, per livestock unit day. In contrast, increased supplies of nitrogen fertilizers to match the increased productivity potential may increase nitrate leaching.