Abstract
Introduction: Studies have investigated the rumen microbiome composition and functions to improve ruminant agriculture and its environmental impacts. Yet, sample collection for rumen microbiome analysis can be difficult and invasive, hindering the ability to sample large animal populations. Studies have proposed using oral swabs as an alternative to rumen sample collection. Here, we investigated the potential of using the oral bacterial community as a proxy for the rumen bacterial community during the cattle production cycle. Methods: We investigated the development of the bovine rumen and oral bacterial communities using longitudinal sampling and the applicability of using the oral to predict host phenotypes. To this end, we utilized 16S rRNA gene sequencing to characterize and compare the rumen and oral bacterial community composition over multiple time points using amplicon sequence variants (ASVs) in a beef cattle population of 166 animals. Additionally, host phenotype of weaning weight was predicted using the Bayesian ridge regression model to evaluate the applicability of using the oral bacterial community for phenotype prediction. Results: Our results identified the rumen and oral bacterial communities to have different trajectories of assembly. The proportion of Proteobacteria and Actinobacteriota was higher (p < 0.0001) in the oral samples. Whereas rumen samples had greater abundance of members of the phyla Bacteroidota, Firmicutes, Verrucomicrobiota, Fibrobacterota, and Spirochaetota. The investigation of the oral and rumen bacterial community establishment demonstrated considerable dynamism, where diet and age-related factors to contribute toward bacterial colonization through introduction of new species and the proliferation of early colonizers. Finally, a Bayesian ridge regression model was developed to estimate weaning weight using the centered and scaled log-transformed relative abundance of ASVs. The proportion of variation explained in weaning weight by the oral and rumen bacterial communities were 30 and 37%, respectively. Discussion: Results from this study suggest that oral and rumen bacterial communities are distinctive, and the oral bacterial community may not serve as a good proxy for the rumen bacterial community even in adult animals with a well-established microbiome. However, the oral bacterial community may serve as a proxy for phenotypic traits of interest in beef cattle.