Competitive exclusion and the No-Interaction model operate simultaneously in microcosm plant communities

Question: Is the relation between productivity and species richness due to an increase in plant size and hence a reduced plant density?Location: Glasshouse experiment.Methods: Productivity was manipulated with fertilizer and irrigation in a microcosm experiment. The 'sampling effect' was removed using rarefaction to a common density of individual plants per pot.Results: Fertility increased community biomass towards an asymptotic maximum, and reduced the light passing through plant canopies towards an asymptotic minimum. As biomass increased, so did species richness. However, this did not seem to be a direct effect of productivity on species richness, but rather one mediated by plant density, since: (1) the richness/ density relation was stronger than the richness/biomass one; (2) adding biomass to the richness/density regression did not increase its predictivity; (3) the richness/biomass relation was removed by rarefaction to 200 individuals per pot. It is therefore concluded that the richness/biomass relation observed was due to the sampling effect. Rarefaction to a small number of plants gave a quite different trend: lower richness at high biomass. This seems to be due to an increased number of subordinate species at high community biomass, and a more uneven distribution of abundance.Conclusion: The Competitive Exclusion and No-Interaction hypotheses have been seen as alternatives. We suggest that they can operate simultaneously.