In natural vegetation~ individuals of different species are more often than not distributed contagiously and arranged as patches. In experimental studies on interference between plant species, this aspect has been neglected.In the experimental approach presented here the vegetation (seven plant species characteristic of woodland clearings) consists of a mosaic of hexagons of 16 m 2 each, in such a way that each monospecific plot of the honeycomb was surrounded by six hexagons planted with other species. In addition, there were separate hexagons without adjacent neighbours. In this way the fate of plants in plots, with and without the influence of vegetatively expanding neighbours could be compared.Ecological advantages of the method applied are a) a high degree of similarity with natural patterns of vegetation, b) the opportunity for the different species to consolidate their own microhabitat, which might have repercussions with regard to succession, c) the comparability of different monocultures with the corresponding units in the mixture.The species specific plant growth (and concurrent capture of nutrients) played an important role in the mechanism of interference (and subsequent succession). Allelopathic effects cannot be excluded.