Segment-Specific Gut Microbiome and Bile Acid Profiles in Grazing and Stall-Fed Yaks

The yak is an iconic ruminant of the Qinghai-Tibet Plateau, yet segment-specific variation in its intestinal microbial functional potential and bile acid profiles under different feeding systems remains insufficiently characterized. Six healthy adult male yaks with similar body weights (320 ± 30 kg) were assigned to grazing (G) or stall-feeding (S) systems, with three animals per group, for a 90-day trial comprising a 10-day adaptation period and an 80-day formal experimental period. The individual yak was considered the experimental unit, and intestinal segments sampled from the same animal were treated as repeated observations. Liver tissue and digesta from the duodenum, ileum, cecum, and colon were analyzed using targeted bile acid metabolomics and shotgun metagenomics. Principal coordinate analysis based on Bray-Curtis dissimilarities showed segment-associated clustering of microbial communities, with PCo1 and PCo2 explaining 65.5% and 18.9% of the total variation, respectively. ANOSIM identified a significant intestinal-segment effect on microbial community composition (R = 0.2208, BH-FDR = 0.0144), whereas the overall feeding-system effect was not significant (R = 0.3747, BH-FDR = 0.1200). No statistically significant feeding-system differences were detected in Shannon, Simpson, Chao1, or ACE indices within individual intestinal segments (BH-FDR ≥ 0.800), and PERMDISP detected no significant differences in within-group dispersion (BH-FDR ≥ 0.1682). Bacillota and Bacteroidota were the dominant phyla. Descriptive functional profiling showed higher mean ileal abundances of GH2 (0.0035 vs. 0.0026), GH3 (0.0029 vs. 0.0024), and GH43 (0.0021 vs. 0.0013) in grazing yaks, whereas the starch-associated GH13 family showed its highest mean abundance in the colon of stall-fed yaks. These metagenomic patterns represent predicted genomic functional potential rather than gene expression, enzyme activity, or metabolic flux. Cecal total bile acid concentration showed a nominal between-group difference (unadjusted Welch's p = 0.0109), but this difference did not remain significant after correction across the five anatomical sites (BH-FDR = 0.0545). In the colon, stall-fed yaks had a lower conjugated-to-unconjugated bile acid ratio and a higher secondary-to-primary bile acid ratio than grazing yaks (BH-FDR < 0.05). Feeding-system-associated descriptive patterns were observed in predicted microbial functional profiles, whereas statistically supported between-group differences were limited mainly to selected colonic bile acid ratios. Given the limited animal-level replication, these findings should be considered exploratory.