Species identity and negative density dependence lead to convergence in designed plant mixtures of twelve species

Abstract

Competition effects on community development are difficult to quantify in species-rich plant communities due to the complexity of possible interactions. We used multispecies mixtures to investigate how species identity and competitive interactions influence the development of plant communities. Given the same set of species with differing initial abundance in various communities, we tested whether communities would become more similar (converge) or dissimilar (diverge) over time depending on the relative importance of species identity and competition. Twenty-four experimental communities were established by planting seedlings of twelve wetland species at different relative abundances and absolute densities. The development of the communities was monitored over three years, and yearly changes in biomass were modelled as a linear function of the species biomass at the start of each period. After three years, a clear dominance structure had developed, with four species making up 80% of the aboveground biomass. In all three years, community dynamics was driven by differences in relative growth rates among the species (i.e. an effect of species identity). However, in the second and third years negative density dependence was also important, with changes in the relative abundance of the most abundant species being negatively related to their biomass at the start of the period. Multiple species interactions – though generally weaker than effects of species identity and intraspecific competition – became increasingly important and also contributed to the dominance pattern. It is concluded that species identity and negative density dependence of the dominant species were the most important factors causing the experimental plant communities to converge. We suggest that model systems composed of several species offer a useful method for investigating the influence of functional traits upon community dynamics.