Plant-soil feedback persists beyond host death to shape density-dependent plant competition

Background and AimsPlant-soil feedback (PSF) shapes plant competition, yet classic PSF experiments often overlook the density dependence and temporal complexities of PSFs, especially after host death. We ask whether microbial effects vary after host death and mediate density dependence in seedling competition. MethodsWe combined forest inventory data from a subtropical forest with a density-gradient greenhouse experiment to evaluate seedling competition among two tree species with different mycorrhizal associations (ectomycorrhizal versus arbuscular mycorrhizal). In 2023, we collected soil inocula from living and dead trees (died between 2014 and 2019) with unsterilized and sterilized treatments. We applied invasion analysis to infer seedling competitive outcomes and used bootstrapping to evaluate uncertainty. ResultsSoil microbial communities shaped seedling competition, favoring the ectomycorrhizal species. Using soils collected from living versus dead hosts as live inocula, we found similar plant competitive outcomes, indicating that PSF persists for over 4-9 years following host death. In contrast, when using sterilized inoculum, we found shifts in competition after host death, suggesting underlying abiotic changes which might masked by microbial effects. Moreover, we found significant evidence that soil microbes can mediate nonlinear density dependence in seedling competition. ConclusionWe provide experimental evidence of persistent microbial legacies that benefit ectomycorrhizal tree species in a subtropical forest. Our study demonstrates how integrating field-based census data with density-gradient experiments and explicit uncertainty estimation can better capture the temporal dimensions, complexities of density dependence, and uncertainties of PSF.