Plant community legacy effects on nutrient cycling, fungal decomposer communities and decomposition in a temperate grassland.

Abstract Soil legacies mediated by plant species-specific microbial communities are major drivers of plant community dynamics. Most soil legacy studies focus on the role of pathogens and mutualists in driving these processes, while much less is known about plant litter-mediated changes to the soil microbial community. Here, we used an existing plant-soil feedback field experiment in which plant communities with different growth strategies (i.e., fast versus slow) and different proportions of functional groups (grasses versus forbs) were allowed to condition the soil over contrasting temporal scales (i.e., one versus two years) in a natural grassland. In the feedback phase, we removed the existent plant community, and replaced it with a standardized response plant community. We then tested the legacy effects of these different soil conditioning treatments on decomposition processes, nutrient cycling and soil decomposer community composition. Soil legacy effects on decomposition and the soil decomposer community composition were most evident right after the start of the feedback phase, but disappeared soon after the new community established. The soil conditioning time and years since disturbance affected most of the soil functions consistently, while no strong effects of plant functional group and plant growth strategy were found. We conclude that after disturbance, it is recovery time, not soil legacy effects, that is the most important factor driving soil functions.