Deciphering the differences of bacterial communities between high- and low-productive wheat fields using high-throughput sequencing

Abstract

Microbial communities have been demonstrated to be essential for healthy and productive soil ecosystems. However, an understanding of the relationship between soil microbial community and soil productivity levels is remarkably limited. In this study, bulk soil (BS), rhizosphere soil (RS), and root (R) samples from the historical high-productive (H) and low-productive (L) soil types of wheat in Hebei province of China were collected and analyzed by high-throughput sequencing. The study highlighted the richness, diversity, and structure of bacterial communities, along with the correlation networks among different bacterial genera. Significant differences in the bacterial community structure between samples of different soil types were observed. Compared with the low-productive soil type, the bacterial communities of samples from the high-productive soil type possessed high species richness, low species diversity, complex and stable networks, and a higher relative abundance of beneficial microbes, such as Pseudoxanthomonas, unclassified Vicinamibacteraceae, Lysobacter, Massilia, Pseudomonas, and Bacillus. Further analysis indicated that the differences were mainly driven by soil organic matter (SOM), available nitrogen (AN), and electrical conductivity (EC). Overall, the soil bacterial community is an important factor affecting soil health and crop production, which provides a theoretical basis for the targeted regulation of microbes in low-productivity soil types.

Keywords: Illumina HiSeq sequencing; bacterial community; mantel test; network analysis; wheat.

Figure 1

Map of the study sites located in Hebei province, which is in the northeast of China. Eight farmlands with high productivity and nine farmlands with low productivity were included in the study.

Figure 2

Principal component analysis (PCA) of soil properties.

Figure 3

(A) Venn diagram of the composition of bacterial communities in the BS, RS, and R samples. Different colors represented different treatments. The numbers in overlapping and non-overlapping sections referred to the quantity of common ASVs and unique ASVs of samples from different types of soil. (B) Non-Metric Multi-Dimensional Scaling (NMDS) analysis of the high-productive and low-productive soil type groups in the bulk soil, rhizosphere soil, and root samples, respectively. The scatter plot of 8 samples from the high-productive group and 9 samples from the low-productive soil type group represented the bacterial ASV community composition. The distance between points represented the degree of difference based on unweighted UniFrac similarities in each group.

Figure 4

Relative abundances on (A) phylum and (B) genus levels of bacteria of different soil and root samples. Cladogram based on LEfSe analysis (LDA > 3.5) of the (C) bulk soil samples bacterial community and (D) rhizosphere soil samples, showing that bacteria that significantly differ between groups, color-coded in red and green. The classification levels from phylum to genus are organized concentrically, with the innermost circle representing phylum and the outermost representing genus. The yellow circle represents ASVs with no significant difference.

Figure 5

The networks of co-occurring bacterial in the bulk soils, rhizosphere soils, and root samples in both high- and low-productive soil type samples of six groups, (A) HBS, (B) HRS, (C) HR, (D) LBS, (E) LRS, and (F) HR, based on correlation analysis. The co-occurring networks are colored by module. Within each panel, the node size corresponds proportionally to its degree of connectivity.

Figure 6

Correlation-based heatmap method reveals the relation between environmental factors and bacterial genera in the (A) BS, (B) RS, and (C) R. The different colors and intensities were adjusted based on associations among traits.