The invasive forb Centaurea stoebe (spotted knapweed) was first observed in the Pacific Northwest (PNW) in the 1890s and has subsequently spread throughout North America. Despite being the point of entry, C. stoebe distribution is limited in the PNW compared to persistent monodominant stands that have established in the intermountain grasslands of Montana (MT). One difference between PNW and MT is annual rainfall, which can influence plant growth and competitive interactions directly, as well as indirectly via moisture-driven shifts in soil pathogens and mutualists. In a greenhouse study we asked if soil moisture influences the competitive ability of C. stoebe, whether this is mediated by plant-soil feedbacks (PSF) with soil biota, and what ecological mechanisms may be involved. We grew C. stoebe with or without the common native grass Bromus marginatus, in live or sterile soil collected from C. stoebe populations in PNW and MT. The soil was watered to maintain either high (~30% volumetric water content) or low (~10%) moisture. We measured C. stoebebiomass, arbuscular mycorrhizal fungal (AMF) colonization and root lesions. In addition, we assessed potential shifts in the whole fungal (ITS) and AMF (18S) community using MiSeq and investigated potential function (AMF or pathogen) using FunGuild.
Centaurea stoebe was largest in sterile wet soil. Growth was reduced by the addition of soil biota in wet, but not dry, soil. This shift from negative to neutral PSF in drier soils also occurred when C. stoebe was grown in competition. In contrast, B. marginatus exhibited negative PSF regardless of soil moisture. Thus, soil moisture influenced PSF differently for each plant, with consequences for the competiveness of C. stoebe. This may be explained by shifts in fungal communities interacting with the roots of C. stoebe and B. marginatus. Specifically, the increase in competitiveness of C. stoebe with decreasing soil moisture may primarily be driven by AMF, rather than pathogens. AMF colonization was highest in competing C. stoebe in dry soils, whereas few lesions (<10%) were observed across treatments. Additionally, the AMF to pathogen ratio (estimated from ITS-sequences) increased in dry soils for C. stoebe, but not B. marginatus. Our results corroborate previous findings showing that C. stoebe competitiveness increases in the presence of AMF. Furthermore, our data show that differences in soil moisture may influence the abundance, composition and function of soil biota, in particular AMF, which may help explain why C. stoebe is invasive in drier regions.