Rumen phenotypic analysis
The rumen phenotypic results of all the groups were shown in Table 1. There were no significant difference (p > 0.05) in the pH、MCP, papillae length and muscle layers thickness between different groups. The papillae width of lambs in EW25 group was significant lower than other groups (p< 0.05). The result indicated that there was little affection of EW on rumen development of dairy lambs.
Sequences Analyses
After sequencing and filtering low-quality and long-section sequences, 785430 counts were obtained from 14 samples. A total of 988806 valid sequences were obtained after quality control. The length of most sequences is between 400-500bp (99.93% of the total).
A total of 9481 OTUs were obtained using QIIME2. Among them, 9482, 8428, 9105 and 8252 OTUs were obtained from EW7, EW15, EW25 and CON group, respectively. There were 140 OTUs were shared by all four groups (Figure 1A). 2038, 1091, 3115 and 2015 OTUs were unique at EW7, EW15, EW25 and CON group, respectively (Figure 1A). The rarefaction curves of all samples tended to be flat, which indicated that the sequencing data volume is reasonable (Figure 1B).
Alpha diversity analysis
As shown in Figure 2, the Chao1, observed species, Shannon of group EW25 were significantly higher than group EW7 and EW15 (p<0.05). There was no significant difference of Simpson among all group (p>0.05). It showed that the species richness of the EW25 group was higher than EW7 and EW15 groups, while there was no significant difference in the abundance of microbial species between the CON and EW groups. The result indicated that there was no distinct affection of EW on rumen microbial species abundance of dairy lambs.
Beta diversity analysis
ANOSIM similarity analysis showed that the test was reliable (p=0.00041) and the difference between groups was higher than the difference within group (R=0.476608) (Figure 3A).
The Principal Coordinates Analysis (PCoA) analysis based on the unweighted unifrac distance matrix of multiple communities was used to judge the similarity of species composition among the samples. The results showed that there was significant difference between EW7 and EW25 groups (p<0.05), the rest showed no significant difference (p>0.05) (Figure 3B).
The cluster analysis showed that the species composition and richness of different groups were similarly, just there was a difference between CON and EW15 groups (Figure 3C).
Bacteria composition analysis
A total of 22 phyla were detected in this study. Among them, there were 9 core flora with relative abundance greater than 0.1%. The distribution of the dominant microbial flora at the phylum level was shown in Figure 4A. The top six common dominant bacterial groups were Firmicutes, Bacteroidetes, Fibrobacteres, Spirochaetes, Proteobacteria and Actinobacteria, of which the relative abundances were all greater than 1%. The 30 most abundant microbial species at the genus level were shown in Figure 4B. The Prevotella and Fibrobacter were dominant genera in four groups.
At tenericutes level, the relative abundance of EW7 group was significantly higher than other three groups (p<0.05) (Figure 5A). The relative abundance of Planctomycetes of EW25 and EW15 groups were higher than other two groups (p<0.05) (Figure 5B). The relative abundance of Fibrobacteres of EW7 group was lower than CON group (P<0.05) (Figure 5C).
At genus level, the relative abundance of Acidaminococcus of EW7 group was significantly higher than EW25 group (p<0.05) (Figure 6A). The relative abundance of Fibrobacter of EW7 group was significantly lower than CON group (p<0.05) (Figure 6B).
LEfSe analysis of intergroup samples
Based on the OTUs level, the specific microorganisms between the 4 groups were summarized through LEfSe analysis and the filter value of LDA Score is 4. Most of bacteria were non-specific and a few statistically and biologically biomarkers belonged to specific bacterial clades (Figure7A). The LDA histogram showed that only 7 specific bacteria genera were significantly enriched in the EW7 group includes Succinivibrionaceae and Aeromonadales, 5 specific bacteria genera were significantly enriched in CON group includes Fibrobacteraceae and Fibrobacterales (Figure7B).
Relationship between rumen phenotypes and bacterial community
The spearman correlation coefficient was used to study the correlation between rumen phenotypes and bacterial species at the genus level (Figure8). Results showed that pH was positively correlated with Rikenellaceae_RC9, Succiniclasticum, F082_unclassified, etc. (p<0.05), and negatively correlated with Ruminococcaceae_UCG-014, Succiniclasticum, Selenomonas, etc. (p<0.05). The content of MCP was positively correlated with CAG-352 and Ruminococcus_2 (p<0.05). The width of rumen papilla was negatively correlated with non-classifiable bacteria (p<0.05). The thickness of rumen muscle layer was positively correlated with Sharpea (p<0.05).