Game theory provides no explanation for seed size variation in grasslands.

Game theoretical models have been suggested to explain the maintenance of a remarkable variation in seed size across species in most types of vegetation. According to these models, which are based on the existence of a trade-off between seed size and seed number; smaller-seeded species can invade any species mixture due to their numerical advantage, and larger-seeded species can invade any species mixture due to their competitive superiority over smaller-seeded species during recruitment. However, till now, there is very limited evidence for seed size effects on recruitment interactions among different species. An experiment was designed using 16 species in Swedish grasslands, varying 384-fold in seed size. Species were sown pairwise and alone in disturbed versus undisturbed small plots in grazed versus ungrazed grassland, and the resulting recruitment was recorded. Seedling densities in the experimental plots were within the natural range. Both disturbance and grazing had a positive effect on recruitment. Seed size did not affect recruitment except from a tendency that increasing seed size made recruitment less dependent on disturbance. Recruitment of smaller-seeded species was not affected by the presence of larger-seeded species. Larger-seeded species did not generally win in direct contest. These results suggest that game theoretical models do not explain maintenance of seed size variation across species in these grasslands. Alternative explanations for seed size variation are that either small-scale heterogeneity provide conditions favouring a range of different seed sizes or other attributes than seed size effectively determine recruitment.