Pastures provide a good model system to study intergenotypic plant interactions because most pastures are mixtures of several sown genotypes and species and the conditions necessary for genetic change are present : (i) an extremely variable biotic and abiotic environment at the local scale, (ii) wide genetic variation, (iii) intense competition so that strong selection is likely, and (iv) long term continuity enables selection to be continuous and cumulative . These together provide a system in which (v) the theoretical outcome of intergenotypic interactions may be readily tested and maximizes the probability of detecting patterns . Interactions between the species will have both ecological and evolutionary consequences, both of which may be of interest to plant breeders . These patterns are detectable at different scales but particular attention will be focused at the level of the individual genotype and especially with mixtures of the grass Lolium perenne and the legume Trifolium repens . Competition experiments between L . perenne and T repens are plentiful, but where specific hypotheses about adaptation at the genotype level have been tested, in all cases T. repens grows best when planted with its natural L . perenne neighbor -a reciprocal effect in the L . perenne has only been reported once . This specific T. repens -L . perenne neighbor recognition may be mediated through soil microorganisms, particularly Rhizobium leguminosarum biovar trifolii and Bacillus polymyxa .