Small-scale genotypic richness stabilizes plot biomass and increases phenotypic variance in the invasive grass Phalaris arundinacea

Aims We aim to understand how small-scale genotypic richness and gen-otypic interactions influence the biomass and potential invasiveness of the invasive grass,Phalaris arundinacea under two different dis-turbance treatments:intact plots and disturbed plots,where all the native vegetation has been removed.Specifically,we address the following questions(i)Does genotypic richness increase biomass production?(ii)Do genotypic interactions promote or reduce bio-mass production?(iii)Does the effect of genotypic richness and gen-otypic interactions differ in different disturbance treatments?Finally(iv)Is phenotypic variation greater as genotypic richness increases?Methods We conducted a 2-year common garden experiment in which we manipulated genotype richness using eight genotypes planted under both intact and disturbed conditions in a wetland in Burlington,Vermont(44°27′23″N,73°11′29″W).The experiment consisted of a randomized complete block design of three blocks,each containing 20 plots(0.5 m2)per disturbed treatment.We calculated total plot biomass and partitioned the net biodiversity effect into three com-ponents:dominance effect,trait-dependent complementarity and trait-independent complementarity.We calculated the phenotypic variance for each different genotype richness treatment under the two disturbance treatments.Important Findings Our results indicate that local genotypic richness does not increase total biomass production of the invasive grass P.arundinacea in either intact or disturbed treatments.However,genotypic interactions underlying the responses showed very different patterns in response to increasing genotypic richness.In the intact treatment,genotypic interactions resulted in the observed biomass being greater than the predicted biomass from monoculture plots(e.g.,overyielding)and this was driven by facilitation.However,facilitation was reduced as genotypic richness increased.In the disturbed treatment,genotypic interactions resulted in underyielding with observed biomass being slightly less than expected from the performance of genotypes in monocultures;however,underyielding was reduced as genotypic richness increased.Thus,in both treatments,higher genotypic rich-ness resulted in plot biomass nearing the additive biomass from indi-vidual monocultures.In general,higher genotypic richness buffered populations against interactions that would have reduced biomass and potentially spread.Phenotypic variance also had contrasting patterns in intact and disturbed treatments.In the intact treatment,phenotypic variance was low across all genotypic richness levels,while in the disturbed treatment,phenotypic variance estimates increased as genotypic richness increased.Thus,under the dis-turbed treatment,plots with higher genotypic richness had a greater potential response to selection.Therefore,limiting the introduction of new genotypes,even if existing genotypes of the invasive species are already present,should be considered a desirable management strategy to limit the invasive behavior of alien species.