Soil moisture and competition determine soil biota effects on invasive Centaurea stoebe

1. Soil biota, such as root-associating fungal mutualists and pathogens, influence plant-plant interactions. Yet, the context-dependency of their effects remains poorly understood, hindering our ability to predict plant invasions. We targeted Centaurea stoebe, an invasive plant which is more successful in arid than mesic North American grasslands. We asked how the abundances of the arbuscular mycorrhizal (AM) fungi and potentially pathogenic fungi are moderated by soil moisture and if shifts in these guilds correspond with C. stoebe growth and competitiveness. 2. We grew C. stoebe plants in pots with every combination of three treatments: soil moisture (wet, dry), soil biota (sterile, live) and competition (none, native grass Bromus marginatus). We measured plant responses and root colonization by AM fungi, identified the relative sequence abundance of AM and potentially pathogenic fungi via guild matches of ITS2 sequences in the database FUNGuild, and examined shifts in AM and potentially pathogenic fungal communities based on the SSU rRNA gene and ITS2, respectively. 3. Soil biota reduced C. stoebe growth in wet but not dry soils, and the presence of the native competitor Bromus marginatus exacerbated these effects. This shift in responsiveness corresponded with relatively fewer AM and more potentially pathogenic fungi in C. stoebe roots grown in wet soils, and vice versa in dry soils. Furthermore, in dry soils, C. stoebe's AM fungal community was strongly dominated by Glomus spp. Conversely, B. marginatus response to soil biota was not affected by soil moisture, and shifts in relative abundance of potentially pathogenic and AM fungi were generally opposite to those observed in C. stoebe. 4. These results suggest that soil biota effects on host plants are species-specific and depend on soil moisture. Furthermore, native plants can promote pathogens in distantly related invasive neighbours, potentially generating biotic resistance. Thus, plant invasions may be mediated by interactions between soil biota, soil moisture, and identity of native competitor. 5. Predicting invasions as the climate continues to change will require surveys and experiments across relevant landscape-level gradients of invasion and environmental factors.