The moderate substitution of Astragalus sinicus returning for chemical fertilizer improves the N cycle function of key ecological bacterial clusters in soil

Abstract

Astragalus sinicus (Chinese milk vetch) is a well-established resource of organic fertilizer widely used in paddy soil to partially replace chemical fertilizers. However, the influence of returning A. sinicus to fields on the soil bacterial community remains poorly understood. Here, we used different amounts of A. sinicus partially replacing chemical fertilizers and investigated the changes in soil physicochemical factors and the soil bacterial community structure responses. Returning A. sinicus to the field significantly increased the soil total nitrogen and available phosphorus content (p < 0.05). Weighted gene correlation network analysis (WGCNA) was applied to detect significant associations between the soil microbiome data and physicochemical factors. Two key ecological bacterial clusters (MEturquoise and MEgreen), mainly containing Acidobacteria, Proteobacteria, and Chloroflexi, were significantly correlated with soil nitrogen (N) levels. A. sinicus partially replacing chemical fertilizers reduced the normalized stochasticity ratio (NST) of rare amplicon sequence variants (ASVs), abundant ASVs, MEturquoise, and MEgreen (p < 0.05). Our results further indicated that a moderate amount of A. sinicus returned to the soil effectively mitigated the trend of reduced relative abundance of N fixation function of key ecological clusters caused by chemical fertilizer. However, a large amount of A. sinicus led to a significant increase in relative abundance of denitrification function and a significant decrease in relative abundance of N fixation function of key ecological clusters. This implies that the moderate substitution of A. sinicus returning for chemical fertilizer improves the N cycling function of key ecological bacterial clusters in soil. From the perspective of the bacterial community in paddy soil, this study provides new insight and a reference on how to find a good balance between the amount of A. sinicus returned to the soil and ecological safety.

Keywords: Astragalus sinicus; bacterial community; ecological cluster; nitrogen cycle function; organic substitution.

Figure 1

Soil bacterial community diversity and structure among different fertilization treatments. (A) Chao1 index of the bacterial community. (B) ACE index of the bacterial community. (C) ß-Diversity of the bacterial community. (D) Correlation analysis between AIP and Chao1. (E) Correlation analysis between AIP and ACE. (F) Correlation analysis between total inorganic N (kg/hm2) and ß-diversity. Different labeled letters indicate significant differences between different treatments according to one-way ANOVA with Duncan’s multiple range tests (p < 0.05).

Figure 2

Relationships between the soil bacterial community and environmental factors. (A) Redundancy analysis between soil microorganisms and environmental factors. (B) Spearman correlation heatmaps of environmental factors and bacterial diversity, dominant genera, and bacterial community assembly. Red represents a positive correlation, and blue represents a negative correlation.

Figure 3

Ecological clusters generated by WGCNA. (A) WGCNA module plot. Dynamic tree cut represents initial clusters. Module colors represent final clusters. Each branch in the hierarchical tree or each vertical line in color bars represents one gene. Genes not attributed to any module are colored gray. (B) Pearson correlation heatmap between ecological clusters and physicochemical factors. (C) Pearson correlation heatmap between different ecological clusters. (D) Distribution of MEgreen in each treatment. Each color represents one module. (E) Distribution of MEturquoise in each treatment. (F) Network and species composition of MEgreen and MEturquoise.

Figure 4

Normalized stochasticity ratio of all ASVs, abundant ASVs, rare ASVs, MEturquoise, and MEgreen in different treatments. Different labeled letters indicate significant differences between different treatments according to one-way ANOVA with Duncan’s multiple range tests (p < 0.05).

Figure 5

Functional abundance of ecological clusters. (A) Abundance of denitrification function in MEturquoise. (B) Abundance of N fixation function in MEgreen. Different labeled letters indicate significant differences between different treatments according to one-way ANOVA with Duncan’s multiple range tests (p < 0.05).