Microbial communities in crop phyllosphere and root endosphere are more resistant than soil microbiota to fertilization

2021 
Abstract Crop harbors diverse microbial communities that profoundly influence host health and agricultural productivity. Crop habitat is a dynamic and heterogeneous environment, and thus crop microbiota may display spatial organization across different compartments. However, impacts of long-term fertilization, as a common agricultural practice, on the assembly of crop microbiota and their relationships with crop yield and quality remain largely unresolved. Here, we collected sorghum roots and leaves, and rhizosphere and bulk soils, from an eight-year field experiment with multiple fertilization regimes, and characterized sorghum-associated microbiomes by amplicon sequencing of bacterial 16S rRNA gene and fungal ITS region. Rhizosphere and bulk soils harbored significantly higher diversity of bacteria and fungi than phyllosphere and root endophytes, and the microbial community composition significantly differed across the four compartments. Fertilization significantly influenced the diversity and compositions of sorghum-associated microbial communities, but had more pronounced effects on rhizosphere and bulk soil microbiomes through altering the relative abundances of some major microbial biomarkers. Bacterial genera Lysobacter, Acidibacter, Steroidobacter, RB41, and Blastococcus, and fungal genera Fusarium and Guehomyces were the dominant microbial predictors of sorghum yield and protein content. Structural equation models revealed that fertilization had a positive and indirect effect on sorghum yield and protein content through influencing the microbial diversity in rhizosphere and bulk soils. We provide novel evidence that microbial communities in sorghum phyllosphere and root endosphere are more resistant than soil microbiota to long-term fertilization, and soil microbiota are important predictors of sorghum yield and protein content.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    58
    References
    5
    Citations
    NaN
    KQI
    []