Results/Conclusions In both 2007 and 2008, the highest mean stem cover of cropfield weeds was found in the low diversity, half-normal seeding rate plots. The results of the linear mixed model analysis suggests the density of seeds sown was more important in predicting the stem cover of weeds than the diversity of plant species that make up that density, particularly in 2008. When compared against the intercept in the best model, the mean stem cover of weeds in the two treatments with the higher seeding densities decreased significantly (P = 0.0411 and P = 0.0160). There was little edge effect as sampling location within a plot was comparatively less important in explaining the occurrence of invasive plants. These results seem to support the fluctuating resource ability theory, in which a plant community becomes more susceptible to invasion whenever there is an increase in the amount of unused resources. As succession proceeds, more specialized native species from the seeding mixes may become more effective in exploiting resources than generalist invasive species and diversity may become more important in determining the invasion resistance of the plots.