There is concern about the potential impact of climate change on the diversity of the UK flora and the fauna it supports. It is likely that the increasing temperatures, changing rainfall patterns and increased likelihood of extreme events predicted by Global Climate Models (GCMs) will alter the current distribution of indigenous plant species and may provide opportunities for non-indigenous plants to invade. The conventional approach to predicting these changes has been to combine species distribution models (based on current habitat range) with GCMs to model the shift in the bioclimatic envelope. However, the species assembly that ultimately occupies a locality will also depend on local conditions, land management and biotic interactions. All of these drivers are likely to respond to climate change and, if the resulting shift in plant functional diversity is to be predicted, they all need to be included in the analysis within a simulation framework based on climate change scenarios predicted for a selected region. Because an extensive literature on the eco-physiology of weeds and their response to the environment and management already exists, arable plant communities are an ideal model system for addressing this fundamental research challenge. We will combine a process based eco-physiological model of weed population dynamics and competition with a stochastic weather generator to predict the shift in the realised niche of indigenous UK arable plants and the potential for non-indigenous species to invade on a regional scale. In addition, the ability of species to adapt and the potential shift in plant functional diversity will be predicted using data on intra and inter specific variability in plant functional traits. The approaches and principles developed in the proposal will be instrumental in improving the predictive understanding and manipulation of plant communities in a range of habitats under climate change including grassland systems.