Variation in species composition among local communities can be due to stochastic events and differences in the ecological interactions that occur throughout a community’s development. This variation can be important for maintaining species coexistence at a regional spatial scale. Understanding the mechanisms that drive communities along alternative trajectories, thus creating variation in community composition, is important for increasing the success of restoration efforts. Disturbance influences grassland plant community diversity, and the variation of disturbance in time can create differences in community composition. There is also debate about whether functional trait redundancy among species in the regional species pool can lead to compositional differences among local communities. I ask, what are the relative contributions of timing of disturbance and functional trait redundancy of the regional species pool to creating variation in community composition in restored grasslands? I designed a spatially explicit restoration experiment in Kansas grassland, employing factorial manipulations of the timing of disturbance (spring, autumn, spring and autumn, or no disturbance) and the functional trait redundancy of the regional species pool (manipulated by adding seeds of 13 functionally redundant species, 13 functionally diverse species, or no species). Each experimental unit was a 16-m² landscape with 1-m² disturbed patches and global seed addition. Early results show that disturbance and seed addition increased local and regional plant species richness, but trait redundancy did not affect seedling establishment. Disturbance, but not seed addition, increased beta diversity, a measure of compositional variation. The elimination of varying disturbance regimes and acute dispersal limitation appear to impair community assembly in this ecosystem, highlighting the importance of recreating both abiotic and biotic interactions to maintain diversity at several spatial scales.