Factors driving global microbiome composition: Beta
diversity
Overall, the first two PCoAs accounted for almost 40% of all variance
in the Bray-Curtis distances across all samples. The clustering patterns
revealed YIAL as an outlier cross, CHILL as intermediate, and the
remaining stocks clustering more closely together on the axes (Figure
3). Using the overall PERMANOVA model, significant population effects
were found in the overall microbial community structure using
Bray-Curtis distances (p = 0.001, R2 = 0.046)
(Table 3). Pairwise PERMANOVA tests using Bray-Curtis distances showed
that YIAL was statistically different from BQ (p = 0.0028), CAP
(p = 0.005), NIT (p = 0.005), PUNT (p = 0.005),
QUIN (p = 0.0065), and RC (p = 0.005) but not from CHILL
(p > 0.05), and all other pairwise PERMANOVA
comparisons showed a lack of significant pairwise differences
(Supplementary Table 5). In the overall PERMANOVA model, sire effects
were found to have significant effects on the microbial community
(p = 0.002, R2 = 0.25; Table 3). Furthermore,
using a unique PERMANOVA model for each cross, sire effects on
Bray-Curtis distances were found to be significant within CHILL and NIT
(p = 0.001, R2 = 0.434 and p = 0.006,
R2 = 0.308, respectively), but no significant sire
effects were found within other crosses (Table 4). Finally, while pen
effects did not contribute significantly to differences in overall
microbiome community composition for Bray-Curtis distances, a
significant difference was found between replicate pens for RC (p= 0.021 and R2 = 0.262, Tables 4).