Soil feedbacks of plant diversity on soil microbial communities and subsequent plant growth

The influence of plant diversity on belowground ecosystem processes and feedbacks of the latter on plant growth are still largely unexplored. A previous field experiment showed positive effects of plant diversity on diversity and activity of culturable soil bacteria. We examined soil carry-over effects of these experimental communities on subsequent plant growth and the development of soil microbial communities in pots. We inoculated sterile substrate with soil from the field experiment and then sowed three plant species - Dactylis glomerata, Plantago lanceolata and Trifolium pratense - as monocultures or 3-species mixtures into the pots. We measured the phytometers over a period of 14 months. Thereafter, we sampled the soil in 3-species mixtures to analyze microbial communities. The original plant species richness and number of functional groups had positive soil carry-over effects on phytometer germination. Furthermore, in mixtures, species richness also stimulated short-term biomass production of phytometers and catabolic diversity of culturable soil bacteria. The number of plant functional groups had a positive effect on the amount of prokaryotic phospholipid fatty acids (PLFAs) but negative effects on the amount of eukaryotic PLFAs and the activity of culturable soil bacteria. The carry-over effects of plant diversity on phytometer growth disappeared with time. However, specific effects of the presence of particular functional groups or species in the original plant communities remained significant and were sometimes paralleled by effects on soil microbial communities. These results suggest that in the short term, soil carry-over effects of plant diversity are mediated by a general stimulation of soil microbes whereas the longer-term effects of particular plant species are more likely due to compositional shifts in soil microbial communities. Although the mechanisms of soil carry-over effects in biodiversity experiments remain to be determined, our study demonstrates that they exist and probably involve soil microbial communities. This 'memory' of soils may be an important factor in ecosystem development.