Rice bacterial leaf blight drives rhizosphere microbial assembly and function adaptation

Abstract

Our results suggest that rhizosphere bacteria are more sensitive to bacterial leaf blight (BLB) than fungi. BLB infection decreased the diversity of the rhizosphere bacterial community but increased the complexity and size of the rhizosphere microbial community co-occurrence networks. In addition, the relative abundance of the genera Streptomyces, Chitinophaga, Sphingomonas, and Bacillus increased significantly. Finally, these findings contribute to the understanding of plant-microbiome interactions by providing critical insight into the ecological mechanisms by which rhizosphere microbes respond to phyllosphere diseases. In addition, it also lays the foundation and provides data to support the use of plant microbes to promote plant health in sustainable agriculture, providing critical insight into ecological mechanisms.

Keywords: metagenomics; microbiome assembly; rhizosphere microbiome; rice bacterial leaf blight.

Fig 1

Alpha diversity and beta diversity of bacterial (A, C) and fungal (B, D) communities in the rhizosphere of healthy and diseased rice. Alpha diversity is characterized based on the Chao1 index. Significant differences were determined using the Kruskal-Wallis test, where asterisks indicate significant differences (*P < 0.05; **P < 0.01; and ***P < 0.001). NMDS is based on Bray-Curtis matrix analysis. Significant differences in communities were tested using PERMANOVA. NB: Ningbo samples, RA: Ruian samples, YJ: Yongjia samples, JH: Jinhua samples, and TZ: Taizhou samples.

Fig 2

Co-occurrence networks between fungal and bacterial kingdoms in diseased (A) and healthy (B) rhizospheres. (C) Degree values of bacterial and fungal taxa in healthy and diseased networks. The significant difference was determined by non-parametric Kruskal-Wallis test. Diseased (D) and healthy (E) comparison of networks’ node-level topological features (degree and closeness centrality). Node size indicates the network degree. Each node represents an ASV, purple represents bacterial ASV, and yellow represents fungal ASV. Correlations are indicated between nodes (correlation coefficient > 0.7 and the green line indicate a positive correlation; correlation coefficient < −0.7 and the red line indicate a negative correlation).

Fig 3

Taxonomic composition of the rice rhizosphere microbial community. RA of different bacteria (A) and fungi (B) in healthy and diseased. The bacteria or fungi ranked outside the TOP10 were grouped into “Others.” NBD: Ningbo diseased samples, NBH: Ningbo healthy samples. RAD: Ruian diseased samples, RAH: Ruian healthy samples. YJD: Yongjia diseased samples, YJH: Yongjia healthy samples. JHD: Jinhua diseased samples, JHH: Jinhua healthy samples. TZD: Taizhou diseased samples, and TZH: Taizhou healthy samples.

Fig 4

Assembly processes of healthy and diseased rhizosphere microbial communities (bacteria and fungi). The weighted βNTI of bacterial (A) and fungal (B) communities. (C) Relative importance of bacterial and fungal community assembly processes. Spearman correlations between βNTI and changes in environmental variables for bacterial (D) and fungal (E) communities. The gray dotted lines indicate the upper and lower thresholds for βNTI = 2 and −2, respectively. Asterisks indicate significant differences (**P < 0.01) and ns indicates no significant differences.

Fig 5

Random-forest model detects bacterial taxa that accurately predict health and disease. The top 60 biomarkers of importance in the random forest classification model. Using random forest classification, all healthy and diseased rhizosphere microbiomes (94 samples in total) were classified at the ASV level and colored at the phylum level. Biomarker taxa were listed in descending order of importance to model accuracy. (B) Results of the predictions based on the random forest model for healthy and diseased rhizosphere samples were not included in the training set. Each square represents a sample in the test sample. (C) 10-fold cross-validation error as a function of the number of input taxa was used to distinguish between the healthy and diseased root microbiota ordered by importance.

Fig 6

Functional diversity and differential changes in healthy and diseased rhizosphere microbiomes based on KO, CAZy, and COG functions. KO (A), COG (B), and CAZy (C) NMDS ranking of functional genes based on Bray-Curtis distance matrix. Volcano plots show enrichment and depletion patterns of healthy and diseased rhizosphere microbes based on KO (D), COG (E), and CAZy (F). Boxplots show the functional alpha diversity of the microbiome based on KO (G), COG (H), and CAZy (I) functions. Alpha diversity was based on the Shannon index. Asterisks indicate significant differences (*P < 0.05).