Linking structure and function in plant-soil feedbacks during Alliaria petiolata invasion

Background/Question/Methods

Invasive plants have the potential to significantly alter plant communities through indirect changes to soil microbial community structure and function. Plant-soil feedbacks have been observed as functional or compositional shifts in several experimental systems, but few studies have attempted to link changes in microbial community metrics with differences in native plant fitness. Alliaria petiolata (garlic mustard) invasions provide an ideal model system due to this plant’s demonstrated impact on arbuscular mycorrhizal fungi. In this study, we collected soil from paired samples in the root zone of A. petiolata and in areas away from the invader at 15 sites throughout this species’ introduced range. Arbuscular mycorrhizal fungi communities present in these soils were characterized based on culture-independent analysis of 18S ribosomal DNA sequences. Invader impacts on soil fungi communities were assessed based on shifts in taxonomic and phylogenetic diversity or structure. Three native plant species with differing mycorrhizal dependencies were grown in the same soils and relative plant biomass was used as our measure of the functional component of plant-soil feedback. Sensitivities of individual fungal taxa to the invader were estimated using relative abundances in soils near to the A. petiolata versus soils farther away.

Results/Conclusions

Across all the soil samples, 225 virtual taxa were identified from 9 families of the order Glomeromycota. Taxonomic richness in all rarefied samples ranged from 39 to 83 virtual taxa and was not significantly different between soil pairs within sites. However, native plants differed in how biomass responded to soil fungi richness. Within sites, the fungal communities tended to be phylogenetically overdispersed based on the net relatedness index, but this was not affected by proximity of A. petiolata. The composition of arbuscular mycorrhizal fungi was significantly different in soils near A. petiolata compared to soils away from the invader. In particular, among a subset of fungal taxa observed frequently enough to allow analysis, the taxa most sensitive to presence of A. petiolata were all members of the Glomeraceae family, whereas taxa most resistant invader presence spanned multiple families including the Glomeraceae, Archaeosporaceae, Gigasporaceae, and Diversisporaceae. The native plants also differed in how much relative fitness benefit was conferred by the presence of sensitive taxa of arbuscular mycorrhizal fungi. Our results suggest that structural changes in soil microbial community during Alliaria petiolata invasion may alter native plant community composition by differentially favoring certain plant species.