Mechanism of microbial action of the inoculated nitrogen-fixing bacterium for growth promotion and yield enhancement in rice (Oryza sativa L.)

Abstract

The use of nitrogen-fixing bacteria in agriculture is increasingly recognized as a sustainable method to boost crop yields, reduce chemical fertilizer use, and improve soil health. However, the microbial mechanisms by which inoculation with nitrogen-fixing bacteria enhance rice production remain unclear. In this study, rice seedlings were inoculated with the nitrogen-fixing bacterium R3 (Herbaspirillum) at the rhizosphere during the seedling stage in a pot experiment using paddy soil. We investigated the effects of such inoculation on nutrient content in the rhizosphere soil, plant growth, and the nitrogen-fixing microbial communities within the rhizosphere and endorhizosphere. The findings showed that inoculation with the R3 strain considerably increased the amounts of nitrate nitrogen, ammonium nitrogen, and available phosphorus in the rhizosphere by 14.77%, 27.83%, and 22.67%, respectively, in comparison to the control (CK). Additionally, the theoretical yield of rice was enhanced by 8.81% due to this inoculation, primarily through a 10.24% increase in the effective number of rice panicles and a 4.14% increase in the seed setting rate. Further analysis revealed that the structure of the native nitrogen-fixing microbial communities within the rhizosphere and endorhizosphere were altered by inoculation with the R3 strain, significantly increasing the α-diversity of the communities. The relative abundance of key nitrogen-fixing genera such as Ralstonia, Azotobacter, Geobacter, Streptomyces, and Pseudomonas were increased, enhancing the quantity and community stability of the nitrogen-fixing community. Consequently, the nitrogen-fixing capacity and sustained activity of the microbial community in the rhizosphere soil were strengthened. Additionally, the expression levels of the nitrogen absorption and transport-related genes OsNRT1 and OsPTR9 in rice roots were upregulated by inoculation with the R3 strain, potentially contributing to the increased rice yield. Our study has revealed the potential microbial mechanisms through which inoculation with nitrogen-fixing bacteria enhances rice yield. This finding provides a scientific basis for subsequent agricultural practices and is of critical importance for increasing rice production and enhancing the ecosystem services of rice fields.

Keywords: nifH gene; Keystone taxa; Microbiome; Nitrogen-fixing bacterium; Rice yield; α-diversity.

Fig. 1

The α-diversity of the nifH gene microbial community was assessed in the rice rhizosphere soil (a) and roots (b) under each treatment; A principal coordinate analysis was conducted on the nifH gene microbial community in the rice rhizosphere soil (c) and roots (d)

Fig. 2

The composition of the nifH gene microbial community was analyzed for the rice rhizosphere soil (a) and roots (b) at the phylum level. Similarly, analyses were conducted at the genus level in rice rhizosphere soil (d) and roots (e). The Venn diagram displays the number of common and unique OTUs present in the rhizosphere soil (c) and roots (f) for each treatment

Fig. 3

The networks of the nifH gene microbial community for rhizosphere soil, CK (a) and R3 (b) treatments, and rice roots, CK (d) and R3 (e), were analyzed at the phylum level. Different colors are used to represent the various phyla. The size of each node, representing an OTU, is proportional to its degree. A blue line denotes a positive correlation, whereas a red line signifies a negative correlation. The Zi-Pi diagram shows the distribution of OTUs, based on their network topology in rhizosphere soil (c) and roots (f), to identify key taxa. Thresholds for Zi and Pi, used to classify OTUs, were set at 2.5 and 0.62, respectively. In the rhizosphere soil (g) and root (h), Linear discriminant analysis (LDA) effect size (LEfSe) (LDA > 3.3, P < 0.05) revealed significant (pink, green, blue, purple) and non-significant (yellow) discriminant classification nodes. The diameter of each circle is proportional to the associated number of taxa. Each concentric ring represents a classification level, progressing from the center to the periphery: domain, phylum, class, order, family, and genus

Fig. 4

The relative expression levels of OsNRT1 and OsPTR9 genes were analyzed in different parts of rice plants. The P value signifies the level of significance between two groups of data; P < 0.05 is considered significant, while P < 0.01 is considered extremely significant

Fig. 5

A correlation heat map was created to depict the associations between the soil environmental factors and nitrogen-fixing bacterial groups in rice rhizosphere soil (a) and roots (b). A Pearson’s correlation matrix, illustrated by the square diagram, depicts the associations between soil factors. Mantel test (c) was used to detect relationships between nitrogen-fixing bacterial communities in rice rhizosphere soil, roots, and various environmental factors and rice-related indicators. EPN: Effective panicle number, AGNPP: Average grain number per panicle, TGW: Thousand grain weight, SSR: Seed setting rate. The Pearson correlation coefficient (r) is represented by hue. Asterisks in the grid and square plots indicate significant Pearson correlation: *, P < 0.05; **, P < 0.01; ***, P < 0.001

Fig. 6

Key taxa predicted to influence nitrate nitrogen content (a) and ammonium nitrogen content (b) in rice inter-root soils were identified through random forest analysis. The mean square error (MSE) percentage increase of the variables was used to determine the significance of these predictors; larger MSE% values indicate greater importance. *: P < 0.05; **: P < 0.01

Fig. 7

Cascading relationships among biological and soil factors with respect to rice yield were examined. The main pathways of influence of latent variables on rice yield were revealed by the partial least squares path model (a) and the standardized effects of factors (b). In the model, each long box corresponds to a latent variable, whereas the parameter within each long box denotes an explicit variable. Larger pathway coefficients are indicated by wider arrow lines, while black and red lines denote positive and negative effects, respectively. *: P < 0.05; **: P < 0.01