Plant-soil feedbacks are shaped by microbial legacies that plants leave in the soil. We tested the persistence of these legacies after subsequent colonization by the same or other plant species using 6 typical grassland plant species. Soil fungal legacies were detectable for months, but the current plant effect on fungi amplified in time. By contrast, in bacterial communities, legacies faded away rapidly and bacteria communities were influenced strongly by the current plant. However, both fungal and bacterial legacies were conserved inside the roots of the current plant species and their composition significantly correlated with plant growth. Hence, microbial soil legacies present at the time of plant establishment play a vital role in shaping plant growth even when these legacies have faded away in the soil due the growth of the current plant species. We conclude that soil microbiome legacies are reversible and versatile, but that they can create plant-soil feedbacks via altering the endophytic community acquired during early ontogeny.

Soil legacy effects are commonly highlighted as drivers of plant community dynamics and species co-existence. However, experimental evidence for soil legacy effects of conditioning plant communities on responding plant communities under natural conditions is lacking. We conditioned 192 grassland plots using six different plant communities with different ratios of grasses and forbs and for different durations. Soil microbial legacies were evident for soil fungi, but not for soil bacteria, while soil abiotic parameters did not significantly change in response to conditioning. The soil legacies affected the composition of the succeeding vegetation. Plant communities with different ratios of grasses and forbs left soil legacies that negatively affected succeeding plants of the same functional type. We conclude that fungal-mediated soil legacy effects play a significant role in vegetation assembly of natural plant communities. and robin.heinen@tum.de † These authors contributed equally to this work.

Our findings highlight how soil fungal and bacterial communities respond to time, season, and plant species identity. We found that succession shapes the soil bacterial community, while plant species and the type of plant species that grows in the soil drive the assembly of soil fungal communities. Future research on the effects of plants on soil microbes should take into consideration the relative roles of both time and plant growth on creating soil legacies that impact future plants growing in the soil. Understanding the temporal (in)stability of microbial communities in soils will be crucial for predicting soil microbial composition and functioning, especially as plant species compositions will shift with global climatic changes and land-use alterations. As fungal and bacterial communities respond to different environmental cues, our study also highlights that the selection of study organisms to answer specific ecological questions is not trivial and that the timing of sampling can greatly affect the conclusions made from these studies.