Residual feed intake phenotype modulates growth performance and ruminal microbiome responses to a Bacillus-based direct-fed microbial in beef steers

Technical Summary: Residual Feed Intake Phenotype Modulates Growth Performance and Ruminal Microbiome Responses to a Bacillus-Based Direct-Fed Microbial in Beef Steers

Problem Statement
Direct-fed microbials (DFMs), particularly Bacillus-based additives, are widely used to enhance ruminant health and productivity by modulating gastrointestinal microbial communities. However, the efficacy of these additives in beef cattle is inconsistent, with responses varying in both magnitude and direction. While previous research attributes this variability to factors such as strain differences, dosage, and diet composition, comparatively little attention has been paid to host-related determinants. Specifically, it remains unclear whether an animal's inherent feed efficiency phenotype, measured as Residual Feed Intake (RFI), influences the response to DFM supplementation. Negative-RFI cattle (more feed-efficient) possess distinct ruminal microbial communities and metabolic profiles compared to positive-RFI cattle (less feed-efficient), suggesting that the ruminal environment may condition the efficacy of microbial additives. This study addresses the gap in knowledge regarding whether RFI phenotype modulates growth performance and ruminal microbiome responses to Bacillus-based supplementation.

Methodology
The study utilized a 2 × 2 factorial experimental design involving 40 crossbred (Angus × Hereford) beef steers selected from an initial pool of 108 based on divergent RFI phenotypes.

• Phenotype Selection: Following a 49-day RFI determination period, 20 negative-RFI steers (mean RFI = −1.66 kg/d; more efficient) and 20 positive-RFI steers (mean RFI = 1.62 kg/d; less efficient) were selected.
• Treatments: Steers were assigned to one of four groups: Control diet (CON) or Bacillus-supplemented diet (BAC) within each RFI phenotype. The Bacillus-based DFM (Papillon, Easton, MD) was administered at 14 g/steer/day (approx. 1.5 × 10⁸ CFU/g) mixed into a corn silage-based total mixed ration (TMR).
• Duration: The feeding trial lasted 70 days.
• Measurements:
  • Growth Performance: Individual dry matter intake (DMI) was recorded continuously using GrowSafe nodes. Body weights (BW) were measured at days 0, 14, 42, and 70 to calculate Average Daily Gain (ADG) and Feed:Gain ratios.
  • Microbiome Analysis: Rumen fluid was collected on day 70. Microbial DNA was extracted, and the V3–V4 regions of the 16S rRNA gene were sequenced on an Illumina MiSeq platform.
  • Bioinformatics: Sequences were processed using QIIME 2 and DADA2 to generate Amplicon Sequence Variants (ASVs). Taxonomic classification was performed against the SILVA 138 database. Data were analyzed using MicrobiomeAnalyst for diversity metrics (Chao1 richness, Bray–Curtis beta diversity) and SAS for statistical modeling of relative abundances.
• Statistical Analysis: Growth and microbiome data were analyzed using mixed models with diet, RFI phenotype, and their interaction as fixed effects. Significance was set at P ≤ 0.05, with tendencies noted at 0.05 < P ≤ 0.10.

Key Results

• Growth Performance: A significant diet × RFI interaction was observed for final BW and ADG (P = 0.02). In negative-RFI steers, Bacillus supplementation significantly increased final BW (489 kg vs. 467 kg) and ADG (2.19 kg/d vs. 2.05 kg/d) compared to the control. Conversely, no performance improvements were detected in positive-RFI steers receiving the additive. DMI was influenced by RFI phenotype (P = 0.03) but not by diet or the interaction.
• Microbiome Diversity:
  • Alpha Diversity: Chao1 richness exhibited a diet × RFI interaction (P = 0.05). Supplementation increased microbial richness in negative-RFI steers but had no effect on positive-RFI steers.
  • Beta Diversity: Community composition (Bray–Curtis dissimilarity) showed a diet × RFI interaction (P = 0.01). Significant differences between control and supplemented diets were found in positive-RFI steers (P = 0.04), whereas no difference was detected in negative-RFI steers (P = 0.31).
• Taxonomic Shifts:
  • Phylum Level: Significant interactions were detected for Bacteroidota and Firmicutes. In negative-RFI steers, supplementation increased Bacteroidota abundance (32.3% vs. 28.6%), whereas Firmicutes abundance did not increase (45.1% vs. 48.1% in BAC− vs. CON−). In positive-RFI steers, supplementation decreased Bacteroidota (24.7% vs. 32.3%) and increased Firmicutes (50.5% vs. 45.9%).
  • Genus Level: Interactions were significant for Prevotella, Lachnospiraceae_NK3A20_group, and Ruminococcus. Bacillus supplementation increased Prevotella abundance in negative-RFI steers (23.5% vs. 21.0%) but decreased it in positive-RFI steers. Conversely, Lachnospiraceae_NK3A20_group and Ruminococcus increased in positive-RFI steers but not in negative-RFI steers.
  • Main Effects: Supplementation increased the relative abundance of Acetitomaculum and Desulfobacterota across phenotypes, while decreasing Ruminobacter.

Significance and Claims
The authors conclude that the efficacy of Bacillus-based DFMs is not uniform but is strongly dependent on the host's feed efficiency phenotype. The study demonstrates that more feed-efficient (negative-RFI) cattle benefit from supplementation with improved growth performance and a microbial shift toward propionate-producing taxa (Bacteroidota and Prevotella). In contrast, less efficient (positive-RFI) cattle exhibited detectable microbial restructuring (enrichment of acetate/butyrate-associated taxa) without corresponding performance gains.

The paper claims that these findings support the development of precision supplementation strategies based on feed-efficiency classification. It emphasizes that taxonomic metrics alone are insufficient to predict DFM efficacy, as functional outcomes (growth) and compositional changes can diverge, particularly in positive-RFI animals. The authors note that while the study identifies phenotype-dependent associations, the correlational nature of the single time-point data precludes establishing causality, and future work incorporating metabolomics (e.g., VFA profiling) is necessary to fully resolve the underlying mechanisms.