Abstract
Plant–soil feedbacks are thought to mediate outcomes of plant competition through microbially driven positive or negative feedback loops. Plant–microbe interactions are known to depend on the underlying environmental context, yet most efforts to understand how plant–soil feedbacks mediate species coexistence have not considered these context dependencies. Here, we use modern coexistence theory to assess how this environmental context‐dependence of plant–microbe interactions might influence plant coexistence, and through which species interactions environmental context‐dependencies are most likely to modify coexistence outcomes. First, we found that the component species interactions in a plant–soil feedback model vary in their relative effects, with microbes impacting fitness ratios more so than direct plant–plant interactions. We also found that linear effects of the environment on plant–soil feedbacks result in nonlinear changes in niche differences and fitness ratios, which can result in altered predictions for coexistence, shifting species pairs from regions of coexistence to competitive exclusion or even priority effects. Finally, we extended our model to simulate how environmental dependence of plant–soil feedbacks in a well‐studied invasive species, garlic mustard ( Alliaria petiolata ), may help explain invasion dynamics across environmental regions. We also discuss paths forward in empirically parameterizing context‐dependent coexistence models using greenhouse experiments. This work builds on recent efforts to integrate plant–soil feedback theory with modern coexistence theory, suggesting that context‐dependencies of plant–microbe interactions may more strongly alter fitness ratios compared to plant–plant interactions. Our results highlight how context‐dependent plant–soil feedbacks may help explain spatial and temporal variation in plant community dynamics and suggest that the underlying mechanism of these dynamics depends on the relative sensitivity of particular species interactions to the environmental gradient.