Changes in plant-soil feedback mechanisms have been proposed to explain the success of non-native plants, including the disruption of mutualisms between natives and mycorrhizal fungi. Garlic mustard (Alliaria petiolata) suppresses mycorrhizal fungi in experimental conditions and is associated with distinct microbial communities in its invaded range, but it is unknown if negative plant-soil feedbacks such as this can explain plant community change in invaded systems.
We studied associations of garlic mustard with understory plant communities and soil biota in eight different forest parcels in the northeastern USA. Densities of all plant species were surveyed, soil pH and total C and N were quantified, and OTU abundance and richness of soil fungal functional groups were determined using Illumina sequencing, on study plots that varied with garlic mustard presence (invaded vs. non-invaded). Plant species diversity responses to garlic mustard presence were tested using generalized linear models, and community composition responses to garlic mustard density were examined using PERMANOVA. Associations among native and non-native plant species, soil fungal functional groups, and soil abiotic characteristics were examined using structural equation models (SEMs).
Abundance and composition of native plants varied significantly with garlic mustard density. Plant Shannon diversity was higher in plots with than without garlic mustard, which was driven by high abundances of other non-native species that co-occurred with garlic mustard. Plant-soil associations were also different between invaded and non-invaded plots. Mycorrhizal, saprophytic, and pathogenic fungi were significant components of the non-invaded model, but mycorrhizae were not important in invaded plots. Several non-native plants other than garlic mustard significantly influenced the model in invaded plots, where they increased saprophyte and decreased pathogen occurrence and had a direct negative effect on common native plant densities. Soil C and N were positively associated with native plant densities in non-invaded plots, but were negatively affected in invaded plots.
Our results indicate that presence of garlic mustard may be enough to significantly disrupt plant-mycorrhizal mutualisms regardless of density, but that densities of other non-native plants have a stronger influence than garlic mustard density on plant-soil feedbacks. These datasets provide an initial framework for understanding how plant invasions can alter native forest understory vegetation via changes in microbial functional diversity, and set the stage for restoration planning and future research.