Abstract
1. Plant soil legacies consisting of species-specific microbial communities are hypothesized to play a critical, structuring role in plant species co-existence processes. Plant species are thought to perform worse on soil conditioned by the same species compared to soil of other species, which serves as a self-limitation mechanism and averts mono-dominance of strong competitors. Here we test in a multi-species community setting, whether root colonisation and resource utilisation of soil patches with distinct soil legacies, are consistent with this hypothesis. 2. We grew eight grassland species together in an outdoor mesocosm setup in unconditioned soil and created soil patches in these communities conditioned by one of four plant species, or a soil mixture of all four. During two subsequent growing seasons, we tested the effect of these conditioned soil patches on belowground root colonisation into the patches of each surrounding plant species using a novel sequencing based approach. In addition, we tested the effect of soil conditioning on local root functioning by injecting tracers into the soil patches and measuring the recovery in aboveground biomass. 3. Against expectations, plant species did not place less roots in own soil patches compared to foreign soil patches, nor did species take up less tracer from own compared to foreign soil patches. Using structural equation modelling, we found that tracer uptake of the plant species was to a varying degree explained by root densities in the various soil patches and by differing soil nutrient availability of the soil patches. We conclude that soil legacy effects are inextricably connected to soil nutrient availability, which needs to be taken into account in plant-soil feedback research to understand the processes that shape plant communities. 4. Synthesis. We found that soil legacy effects in complex, multi-species semi-field conditions did not match expectations based on theory and experiments in controlled conditions. Among the many complicating factors that may modify or even overrule soil legacy effects in semi-field settings, we identified soil nutrient availability as a critical force that may, together with soil biota, shape plant species co-existence processes.
