Positive biodiversity-productivity relationships have been consistently observed in grassland experiments, with biodiversity effects increasing over time. However, the mechanisms behind such increase in biodiversity effects remain unresolved. In the present study we hypothesize that both plant-soil feedback and plant community history independently and interactively drive the increase in biodiversity effects in grassland plant communities. In a large biodiversity experiment in Jena, Germany, we re-established the communities of varying plant species richness and composition, taking advantage of history of the past 8 years to establish the plant and soil communities. We reassembled the same experimental communities with seeds from the plants of the same plot (common plant history) or from the supplier which provided the original seeds. We used unsterilized or sterilized soil taken from the same plots and in the case of sterilized soil either re-inoculated with the original (common soil history) or with the neutral soil community. We hypothesized that biodiversity effects on productivity, based on “community assembly“ and plant-soil interaction would be stronger in re-assembled communities with common plant and soil community history.
Results/Conclusions
The positive biodiversity-productivity relationship was maintained across all soil and plant community treatments. In the absence of soil history, common plant community history enhanced community productivity. Soil community history had contrasting effects across the diversity gradient, indicating a role of generalist pathogens reducing productivity at high diversity. Soil microbial biomass data and TRFLP data also support the effect of different soil and plant community treatments on community productivity across biodiversity. Our results clearly demonstrate that soil and plant community history exert substantial effects on biodiversity-ecosystem function relationships, and in part explain the strengthening of biodiversity-productivity relationships commonly observed in long-term biodiversity experiments.