Dietary energy source and level in lactation diets can profoundly affect milk yield and composition. Such dietary effects on lactation performance are underpinned by alteration of the rumen microbiota, of which bacteria, archaea, fungi, and protozoa may vary differently. However, few studies have examined all the four groups of rumen microbes. This study investigated the effect of both the level and source of dietary energy on rumen bacteria, archaea, fungi, and protozoa in the rumen of lactating dairy cows. A 2 × 2 factorial design resulted in four dietary treatments: low and high dietary energy levels (LE: 1.52-1.53; and HE: 1.71-1.72 Mcal/kg dry matter) and two dietary energy sources (GC: finely ground corn; and SFC: steam-flaked corn). We used a replicated 4 × 4 Latin square design using eight primiparous Chinese Holstein cows with each period lasting for 21 d. The rumen microbiota was analyzed using metataxonomics based on kingdom-specific phylogenetic markers [16S rRNA gene for bacteria and archaea, 18S rRNA gene for protozoa, and internally transcribed spacer 1 (ITS1) for fungi] followed with subsequent functional prediction using PICRUSt2.

Dietary energy level affected the overall rumen multi-kingdom microbiota while little difference was noted between ground corn and steam-flaked corn. Some genera were also affected differently by the four dietary treatments, including genera that had been shown to be correlated with lactation performance or feed efficiency. The co-occurrence patterns among the genera exclusively found for each dietary treatment may suggest possible metabolic interactions specifically affected by the dietary treatment. Some of the major taxa were positively correlated to milk properties and may potentially serve as biomarkers of one or more lactation traits.

The GC resulted in a higher prokaryotic (bacterial and archaeal) species richness and Faith's phylogenetic diversity than SFC. For the eukaryotic (fungi and protozoa) microbiota, the LE diets led to significantly higher values of the above measurements than the HE diets. Among the major classified taxa, 23 genera across all the kingdoms differed in relative abundance between the two dietary energy levels, while only six genera (none being protozoal) were differentially abundant between the two energy sources. Based on prokaryotic amplicon sequence variants (ASVs) from all the samples, overall functional profiles predicted using PICRUSt2 differed significantly between LE and HE but not between the two energy sources. FishTaco analysis identified Ruminococcus and Coprococcus as the taxa potentially contributing to the enriched KEGG pathways for biosynthesis of amino acids and to the metabolisms of pyruvate, glycerophospholipid, and nicotinate and nicotinamide in the rumen of HE-fed cows. The co-occurrence networks were also affected by the dietary treatments, especially the LE and GC diets, resulting in distinct co-occurrence networks. Several microbial genera appeared to be strongly correlated with one or more lactation traits. Conclusions: Dietary energy level affected the overall rumen multi-kingdom microbiota while little difference was noted between ground corn and steam-flaked corn. Some genera were also affected differently by the four dietary treatments, including genera that had been shown to be correlated with lactation performance or feed efficiency. The co-occurrence patterns among the genera exclusively found for each dietary treatment may suggest possible metabolic interactions specifically affected by the dietary treatment. Some of the major taxa were positively correlated to milk properties and may potentially serve as biomarkers of one or more lactation traits.