3.1 Variation of gut bacterial composition due to giant panda diet conversion
After quality filtering, a total of 13,650,999 bacterial 16S rRNA gene sequences were obtained from 168 fecal samples from the diet conversion experiment. The sequences were clustered into 3,027 OTUs at the 97% sequence identity threshold. Both community richness (Chao1 index) and diversity (Shannon index) varied with host diet and significant differences in these values were observed between the three experimental groups (p < 0.05, ANOVA test) (Figure 1a, b). Specifically, gut bacterial diversity increased when transitioning from OMD to MBD and OBD diets, while richness conversely declined. PCoA analysis also indicated that samples from the same diet group clustered together and separately from those of different groups (Figure 1c). Indeed, significant differences in community structure were identified among the three groups based on BC distances (p < 0.05, AMOVA).
Proteobacteria and Firmicutes were the dominant phyla among communities sampled during diet conversion, comprising more than 90.0% of the total sequences (Figure 1d). However, Proteobacteria was the most dominant phylum in OMD (comprising 85.5% of the total sequences) and MBD (57.7%) communities, while Firmicutes was the most dominant phylum in OBD (58.3%) communities. Further, the abundance of Proteobacteria was significantly highest in OMD (85.5%) communities than in other groups and was considerably lower in the OBD (35.1%) communities (non-parametric factorial Kruskal-Wallis sum-rank test, LDA > 4) (Figure 1e). Conversely, Firmicutes abundances increased markedly from the OMD (13.5%) to the OBD (58.3%) communities (Non-parametric factorial Kruskal-Wallis sum-rank test, LDA>4), and were significantly higher in the OBD samples (Figure 1d, e).
The distribution of the 10 most abundant genera in each group (comprising > 80.0% of the total sequences in each group) were further investigated (Figure 1f). The three most abundant genera in the OMD samples were Escherichia-Shigella (80.1%), Streptococcus (7.9%), and Lactobacillus (1.9%). The abundances of Escherichia-Shigella sharply decreased in the OBD communities relative to the OMD communities (Non-parametric factorial Kruskal-Wallis sum-rank test, LDA>4) (Figure 1e, f). The three most abundant genera in the MBD communities were Escherichia-Shigella(43.8%), Streptococcus (16.8%), and Lactobacillus(10.1%). Streptococcus abundances were significantly higher in the MBD group than in other groups (7.9% and 11.4% in OMD and OBD) (Non-parametric factorial Kruskal-Wallis sum-rank test, LDA>4) (Figure 1e, f). Lastly, Pseudomonas(13.4%), Lactobacillus (12.7%), and Clostridium (12.2%) were the three most abundant genera in the OBD communities. In addition,Pseudomonas, Lactobacillus , Clostridium, Enterococcus, Lactococcus, Turicibacter, Acinetobacter , Cetobacterium, andHafnia-Obesumbacteriumabundances also significantly increased when transitioning from the OMD (0.2%, 1.9%, 0.1%, 0.4%, 1.0%, 0.0%, 0.2%, 0.0%, and 0.0%, respectively) to the OBD (13.4%, 12.7%, 12.2%, 7.0%, 4.8%, 3.9%, 3.9%, 3.6% and 2.4%, respectively) communities (Non-parametric factorial Kruskal-Wallis sum-rank test, LDA>4) (Figure 1e, f).