Since botanical diversity in grassland has a tendency to be inversely proportional to grass productivity, efforts at encouraging species rich grassland will incur production penalties. Development of management guidelines to achieve particular levels of biodiversity by evaluation of production losses and conservation benefits is required, together with cost effective ways of creating or enhancing species diversity in semi-natural grassland. This study will determine cost effective methods of creating/enhancing species rich grassland, with particular reference to Lolium agrostis/fescue grassland in upland areas which have previously been fertilised using nitrogen and intensively managed for sheep production. The project will advance understanding of the ecological processes determining botanical diversity in de-intensified swards and in reconstructed swards. In particular, the role of buried seeds, gap size, seed rain, root competition, defoliation and predation will be determined.