Seasonal stability of the rumen microbiome contributes to the adaptation patterns to extreme environmental conditions in grazing yak and cattle

Abstract

Background: The rumen microbiome plays an essential role in maintaining ruminants' growth and performance even under extreme environmental conditions, however, which factors influence rumen microbiome stability when ruminants are reared in such habitats throughout the year is unclear. Hence, the rumen microbiome of yak (less domesticated) and cattle (domesticated) reared on the Qinghai-Tibetan Plateau through the year were assessed to evaluate temporal changes in their composition, function, and stability.

Results: Rumen fermentation characteristics and pH significantly shifted across seasons in both cattle and yak, but the patterns differed between the two ruminant species. Ruminal enzyme activity varied with season, and production of xylanase and cellulase was greater in yak compared to cattle in both fall and winter. The rumen bacterial community varied with season in both yak and cattle, with higher alpha diversity and similarity (beta diversity) in yak than cattle. The diversity indices of eukaryotic community did not change with season in both ruminant species, but higher similarity was observed in yak. In addition, the similarity of rumen microbiome functional community was higher in yak than cattle across seasons. Moreover, yak rumen microbiome encoded more genes (GH2 and GH3) related to cellulose and hemicellulose degradation compared to cattle, and a new enzyme family (GH160) gene involved in oligosaccharides was uniquely detected in yak rumen. The season affected microbiome attenuation and buffering values (stability), with higher buffering value in yak rumen microbiome than cattle. Positive correlations between antimicrobial resistance gene (dfrF) and CAZyme family (GH113) and microbiome stability were identified in yak, but such relationship was negatively correlated in cattle.

Conclusions: The findings of the potential of cellulose degradation, the relationship between rumen microbial stability and the abundance of functional genes varied differently across seasons and between yak and cattle provide insight into the mechanisms that may underpin their divergent adaptation patterns to the harsh climate of the Qinghai-Tibetan Plateau. These results lay a solid foundation for developing strategies to maintain and improve rumen microbiome stability and dig out the potential candidates for manufacturing lignocellulolytic enzymes in the yak rumen to enhance ruminants' performance under extreme environmental conditions.

Keywords: Cattle; Microbiome functionality; Rumen metagenome; Rumen microbiome stability; Yak.

Fig. 1

Experimental design for rumen sample collection from grazing yak (n = 6) and cattle (n = 6) across seasons

Fig. 2

Microbial compositional profiles between yak (n = 6) and cattle (n = 6) at different seasons. Microbial domains composition between cattle (A) and yak (B). Non-metric multidimensional scaling (NMDS) analysis plots based on Bray–Curtis metrics showed bacterial (C) and archaeal (D) communities (at species level) of yak and cattle are different from each other at different seasons

Fig. 3

Temporal changes of rumen microbial alpha diversity between yak (n = 6) and cattle (n = 6) at different seasons. A Bacterial Shannon and Chao1 of yak and cattle at different seasons. B Archaeal Shannon and Chao1 of yak and cattle at different seasons. C Fungal Shannon and Chao1 of yak and cattle at different seasons. D Protozoal Shannon and Chao1 of yak and cattle at different seasons. Differences in data were assessed by non-parametric Kruskal–Wallis test in combination with Dunn’s post-doc test for multiple comparisons, and all P values were corrected by Benjamin-Hochberg algorithm (* 0.05 < P < 0.01, ** 0.01 < P < 0.001, *** P < 0.001)

Fig. 4

Dynamic change in composition of rumen microbiome between yak (n = 6) and cattle (n = 6) at different seasons. Composition of rumen bacteria (A) and archaea (B) at the phylum level (relative abundance > 0.1% in at least in 50% of the at least one timepoint within each species). Significantly different (P < 0.05) bacterial (C), archaeal (D), fungal (E), and protozoal (F) species between cattle and yak across seasons. Significantly different was assessed by non-parametric Kruskal–Wallis test in combination with Dunn’s post-doc test for multiple comparisons

Fig. 5

Temporal changes of rumen functional profiles between yak (n = 6) and cattle (n = 6) at different seasons. Non-metric multidimensional scaling (NMDS) analysis plots based on Bray–Curtis metrics showed KEGG profiles (A, at level 2), CAZyme profiles (B), and ARGs profiles (C) between yak and cattle at different seasons

Fig. 6

Rumen microbiome KEGG pathways function between yak (n = 6) and cattle (n = 6) at different seasons. Composition of KEGG (level 2) at different seasons between cattle (A) and yak (B). Significantly different (P < 0.05) KEGG pathways between cattle (C) and yak (D) across seasons. Significantly different was assessed by non-parametric Kruskal–Wallis test in combination with Dunn’s post-doc test for multiple comparisons

Fig. 7

Rumen microbiome CAZymes between yak (n = 6) and cattle (n = 6) at different seasons. Composition of CAZymes at different seasons between cattle (A) and yak (B). Significantly different (P < 0.05) CAZyme families between cattle (C) and yak (D) across seasons. Significantly different was assessed by non-parametric Kruskal–Wallis test in combination with Dunn’s post-doc test for multiple comparisons

Fig. 8

Rumen microbiome ARGs profile between yak (n = 6) and cattle (n = 6) at different seasons. Composition of ARGs at different seasons between cattle (A) and yak (B). Significantly different (P < 0.05) ARGs between cattle (C) and yak (D) across seasons. Significant difference was assessed by non-parametric Kruskal–Wallis test in combination with Dunn’s post-doc test for multiple comparisons, and the circos plots were generated using circlize R package [128]

Fig. 9

Temporal changes of attenuation and buffering values in rumen microbiota between yak (n = 6) and cattle (n = 6) across seasons. Statistical analysis was determined using repeated measures ANOVA and the Tukey test for multiple comparisons, and P values were corrected by Benjamin-Hochberg algorithm (* 0.05 < P < 0.01, ** 0.01 < P < 0.001, *** P < 0.001)

Fig. 10

The potential relationship among ARGs, CAZymes and stability (Attenuation and Buffering). Correlation analysis was performed by Spearman rank correlation, and only correlations (among ARGs, CAZymes and stability) that were positive or negative at the same time were displayed. A Relationship between ARGs and stability in cattle. B Relationship between ARGs and stability in yak. C Relationship between CAZymes and stability in cattle. D Relationship between CAZymes and stability in yak. The lines in red and blue denote positive and negative correlations, respectively

Fig. 11

The dynamic changes of rumen microbial community and function among different studies. A Bacterial community similarity between the current study (cattle and yak) and other studies. B Archaeal community similarity between the current study (cattle and yak) and other studies. C KEGG pathways similarity between the current study (cattle and yak) and other studies. D CAZyme family similarity between the current study (cattle and yak) and other studies. E The distributions of CAZyme families in different studies. Studies 1–4 indicates that the published studies (* 0.05 < P < 0.01, ** 0.01 < P < 0.001, *** P < 0.001)