Limited effects of dominant species population source on community composition during community assembly

Question To what extent do dominant species population sources and subordinate species pools affect diversity and composition of an assembling grassland community? Location Illinois, USA. Methods Percentage cover of all species were recorded annually in 36 1-m2 quadrats assigned to a factorial combination of dominant species population source (functionally distinct cultivar or non-cultivar seed source) and designed species pool (three levels varying in species identity, but with equal functional group representation and richness) during the first 4yr of community assembly in an experimental grassland restoration. Results Univariate and multivariate analyses showed that individual species abundance, life form and community composition differed significantly among designed species pools, but were not strongly affected by population source of the dominant species (cultivar or non-cultivar). There were fewer C4 species in cultivar plots but only in one of three designed species pools during one of 4yr of community assembly. The number of legume and forb species was higher in cultivar plots, but also only in one of the 4yr of study. Other changes in species richness and abundance were solely related to successional change. Conclusions Non-dominant species introduced to restore plant communities strongly affects plant community composition, and composition can show fidelity to designed species pools. Only marginal or temporary effects of dominant species seed source were observed in the assembling plant community. Thus, we found no strong evidence that the source of dominant species, in this case cultivars compared to local ecotypes, has consequences for community assembly in the early stages of restoration (14yr). The absence of a strong dominant species source effect may be exacerbated by the assembly of diverse plant communities, resulting in a stronger effect of subordinate species seed mixture in restoration.