Formation of high-quality mixed silage from paper mulberry and wheat bran driven by the characteristics of the microbial community

Abstract

Paper mulberry (Broussonetia papyrifera) is a high-quality silage protein feed material that can help address feed shortages and support livestock development. Although some studies have investigated the relationships between microbial communities and silage quality, these relationships and the underlying community assembly processes remain complex, requiring further research to clarify them. Additionally, limited research has explored the relationship between microbial community fermentation functions and silage quality. In this study, we aimed to explore B. papyrifera and wheat bran mixed silage quality driven by the characteristics of the microbial community. After 50 days of silage fermentation, high-quality and low-quality samples were selected from every mixing ratio (90:10, 80:20, and 65:35). The silage chemical composition, lignocellulose degradation enzyme activity, microbial community composition, and potential functions were used to explore the relevance between silage quality and the characteristics of the microbial community. The contents of hemicellulose, neutral detergent fiber, pH, and the activities of endoglucanase and exoglucanase were significantly affected by mixing ratios and silage quality grade. There were higher crude protein content, lignocellulose degrading enzyme activity, and lower pH, lignin, and acid detergent fiber in the mixing of 65:35 (BP65%) samples. The PERMANOVA results showed that mixing ratios had significant impacts on microbial community composition and bacterial fermentation functions. There was a higher bacterial diversity, lower fungal diversity, and better functional potentials for fermentation and lignocellulose degradation in BP65% high-quality silage. The dominant genera were Lactobacillus, Cladosporium, and Wallemia in all samples. The relative abundance of Clostridium, Rhodococcus, Turicibacter, Ralstonia, and Burkholderia was significantly higher in BP65% high-quality samples. There was a higher abundance of Wallemia in the BP65% samples than in other mixing ratios samples. Notably, silage quality showed a close relationship with Lactobacillus, Turicibacter, Romboutsia, Wallemia, and Pichia. In summary, 65:35 was a suitable mixing ratio for B. papyrifera and wheat bran silage, but high-quality silage still required the participation of multiple specific rare microbial taxa. The higher bacterial diversity and specific microbial taxa abundance could be critical for improving B. papyrifera silage quality. We expect that our findings will provide new insights into silage quality driven by the characteristics of the microbial community.

Keywords: fermentation function; lignocellulose degradation; microbial community; mixed ensiling; silage quality.

Figure 1

Chao and Shannon indices of bacterial and fungal communities. (A) Chao index of bacterial community; (B) Shannon index of bacterial community; (C) Chao index of fungal community; and (D) Shannon index of fungal community.

Figure 2

Principal coordinates analysis (PCoA) of bacterial (A) and fungal (B) community compositions.

Figure 3

Bacterial (A) and fungal (B) community structures at the level of genus.

Figure 4

Variations in bacterial community composition of phenotype. (A) Functional prediction of bacterial community FAPROTAX and (B) test for inter-group differences in predictive function.

Figure 5

Bacterial community FAPROTAX functions prediction (A) and functional group difference analysis (B).

Figure 6

Correlation heatmap of microbial taxa and silage quality properties. (A) Bacterial community and lignocellulose content; (B) bacterial community and lignocellulose degradation enzymes; (C) fungal community and lignocellulose content; and (D) fungal community and lignocellulose degradation enzymes. FPA, filter paper cellulase CP, crude protein; CL, cellulose; HC, hemicellulose; LI, lignin; ADF, acid detergent fiber; NDF, neutral detergent fiber. *p < 0.05, **p < 0.01, ***p < 0.001.