Long-Term Organic-Inorganic Fertilization Regimes Alter Bacterial and Fungal Communities and Rice Yields in Paddy Soil

Abstract

Microorganisms are the most abundant and diverse organisms in soils and have important effects on soil fertility. In this study, effects of the long-term fertilization treatments no fertilizer (CK), chemical fertilizer (nitrogen-phosphorus-potassium (NPK)), and organic-inorganic fertilizer (NPK and organic fertilizer (NPKM)) on rice yield and soil bacterial and fungal community diversity, structure, composition, and interaction networks were evaluated. Of the three treatments, the highest rice yield was in NPKM. Bacterial richness was significantly higher in NPKM than in NPK. Fertilization treatment significantly altered β diversity of communities, species composition of bacterial and fungal communities, and structure of soil microbial networks. The most complex bacterial and fungal interaction co-occurrence network with the highest average degree and numbers of edges and nodes was in NPKM. Relative abundance of the plant growth-promoting fungus Trichoderma increased significantly in NPKM compared with CK and NPK. The results of the study indicate that bacterial richness and microbial community member interactions (network complexity) might be suitable indicators of soil biological fertility. This research provides new insights on the effects of different fertilization regimes on responses of soil bacterial and fungal communities and their contributions to crop yield. New indicators such as bacterial richness and complexity of microbial interaction networks are also identified that can be used to evaluate soil biological fertility.

Keywords: co-occurrence network; fertilization regimes; microbial community; rice; soil biological fertility.

Figure 1

Effects of long-term fertilization treatments on rice yield. Different letters indicate significant differences among treatments (Tukey’s post hoc test, p < 0.05).

Figure 2

Effects of long-term fertilization treatments on Chao1 richness, evenness, and Shannon diversity indices of (A–C) bacteria and (D–F) fungi. Different letters indicate significant differences among treatments (Tukey’s post hoc test, p < 0.05).

Figure 3

Unconstrained principal coordinate analysis (PCoA) with weighted UniFrac distance showing effects of long-term fertilization treatments on beta-diversity of soil (A) bacterial and (B) fungal communities.

Figure 4

Effects of long-term fertilization treatments on composition of bacterial and fungal communities. Phylum-level composition of (A) bacterial and (C) fungal communities. Analysis of differences in relative abundances of (B) bacterial phyla in A and (D) fungal phyla in C. Names of the nine most abundant phyla are shown, and other phyla were grouped as “Others.” Different letters indicate significant differences (Tukey’s post hoc test, p < 0.05).

Figure 5

Heat map of relative abundances of genera of bacteria and fungi with positive correlations to rice yield. Relative abundances were Z-score normalized. Genus-level bacteria names begin with “b”, and genus-level fungi names begin with “f”.

Figure 6

Co-occurrence networks of bacterial and fungal communities in (A) control (CK), (B) nitrogen–phosphorus–potassium (NPK), and (C), NPK plus organic fertilizer (NPKM) fertilization treatments. Each node in a network represents a genus of bacteria or fungi. Size of a node represents average relative abundance of a genus of bacteria or fungi. Different colors of nodes provide taxonomic information at the phylum-level. In the legend, names of bacterial phyla begin with “b,” and names of fungal phyla begin with “f.” (D) Common and unique edges in co-occurrence networks between each two of the three treatments. (E) Comparison of sequence numbers of the fungus Trichoderma in different fertilization treatments. (F) Indices related to topological structure of the co-occurrence interaction network of bacterial–fungal communities.

Figure 7

Effects of long-term fertilization treatments on the ratio of Proteobacteria to Acidobacteria. Different letters indicate significant differences (Tukey’s post hoc test, p < 0.05).