Phytomass distribution and species diversity were followed in a chalk grassland with different selective clipping treatments during 4 years. Removal of the dominant Brachypodium[pinnatum] resulted in a sharp increase in forb phytomass and in species number. The dominance-diversity curve changed in shape from geometric to log-normal. Obviously, Brachypodium restricted the growth of most other species. Removal of all species except the dominant showed, in contrast to the niche pre-emption theory, a marked increase in phytomass and reproductive effect of Brachypodium. It seems that other mechanisms can also lead to a geometric distribution. Removal in a type of vegetation without obvious dominants and with a log-normal species distribution revealed that interspecific competition is not an important factor in this vegetation. This suggested a low intensity of species interaction and a large influence of restricting density-independent factors. Finally, to reduce the dominance of Brachypodium we implemented a 4-year mowing experiment. Mowing earlier in the season resulted in a reduced exclusion of species by Brachypodium. This alternative mowing time may be sufficient to restore species diversity in chalk grassland. Experiments are under way to further substantiate this conclusion.