Historical ranching operations in California have transformed much of the native coastal sage scrub (CSS) to nonnative grasslands. One nonnative invasive bunch-grass, Phalaris aquatica, had replaced CSS, and was subsequently removed between 2006–2013 from a 25-acre field site in the Santa Monica Mountains National Recreation Area. In a previous study, we demonstrated that biotic soil legacy effects due to P. aquatica were significantly associated with decreased native plant growth. Our questions in this study: which bacterial taxa are responsible for the difference in native plant growth, and can the inhibition of native plant growth in post-P. aquatica soil be prevented by inoculation with native microbes? To address these questions, three species of native plants (Artemisia californica, Salvia leucophylla, Baccharis pilularis) were grown in pots filled with either sterilized or unsterilized native CSS soil collected near the 25-acre post-P. aquatica site (N=30 pots per treatment). A total of 180 native plants were transplanted from these pots into the 25 acre post-P. aquatica site. During seven months of growth in the field, monthly soil cores were taken for nitrogen and microbial composition analysis (Illumina MiSeq sequencing), plant height and mortality was recorded, and (in the last month) rhizosphere soil was collected.
Results/Conclusions
Native plants grew better with the addition of unsterilized native soil inoculum than with sterilized native soil inoculum. Artemisia californica had a 3:1 mortality ratio for plants grown in sterilized vs. unsterilized native soil, while A. californica and B. pilularis grew significantly taller (p= 0.012, 0.007) in unsterilized vs. sterilized native soil. The increased native plant growth and survival in the post-P. aquatica site occurring after initial growth in unsterilized–but not sterilized–native soil inoculum, suggests that differences in the microbial composition of the soils are causing differences in plant growth. For the first two months of growth, the microbial community composition between soil treatments were the same, but different than intact CSS and post-P. aquatica control soils. Soil invaded by P. aquatica had higher amounts of Actinomycetales and Sphingomonadales than native soil and a lower amount of Rhizobiales and Rubrobacterales. In the last month of growth, rhizosphere soil and roots from destructively sampled plants had different microbial composition between treatments. The differences in microbial communities observed between soil treatments at various taxonomic levels, especially in the last month of the study, indicate that soil restoration via native soil inoculum will improve native plant restoration success in post-invasive-grass soils.