Plant density affects measures of biodiversity effects

Aims:We tested for the effect of final sowing plant density(i.e.density of established seedlings)on the values of biodiversity effects[transgressive overyielding,net effect,complementarity effect(CE)and selection effect(SE),trait-dependent complementarity and dominance effect]in a glasshouse pot experiment.Methods:We conducted a single-season(4 months)glasshouse experiment.Species monocultures and mixtures containing up to four common meadow species from different functional groups were sown and subsequently thinned to five density levels(8–128 individuals per pot,i.e.200–3200 individuals m 2).Community functioning was characterized by yield(both living and dead biomass)of all constituent species.Important Findings:Our results show that plant density(final sowing density in our case,but this finding can be generalized)affects the yields of both monocultures and mixtures.As these and their relationships are the basis for calculation of biodiversity effects,these effects also varied along the density gradient.Net biodiversity effect,CE and SE all increased with density.The net biodiversity effect and the CE switched from negative to quite positive in the four-species mixture.Using Fox’s tripartite partitioning,trait-dependent complementarity was minor in comparison to the dominance effect.One of our experimental species did not follow the density–productivity relationship,called constant final yield(CFY),which was reflected in the biodiversity measures.The shape of the density–productivity relationship for experimental species affects also the values of biodiversity indices,particularly when species do not follow the CFY relationship.According to our data and recent simulation experiments,the values of commonly used biodiversity effects can be rather misleading if a species has,e.g.a unimodal dependence of yield for the density gradient and the density level used in the experiment is higher than the peak density.