Biochemical and Microbial Biomarkers of Feed Efficiency in Black Angus Steers

As the global population is expected to exceed 9 billion people by 2050, finding novel methods of improving food production is imperative. The rumen microbi-ome is critical in ruminant nutrition and contributes to nutrient utilization and feed efficiency in cattle. Therefore, the objective of this study was to interrogate microbial and biochemical factors affecting divergences in feed efficiency in Black Angus steers. Fifty Black Angus steers of 7 months of age, weighing 264 ± 2.7 kg were acclimated to the GrowSafe© feeding system for lOd prior to intake measurement, and fed a step-up receiving diet 14d before receiving a growing ration (11.57% CP and 76.93% TDN DM) with 28 mg monensin/kg DM. Steers were maintained on the diet for 70d. Weekly BW was measured, serum collected, and rumen content was obtained via gastric tubing. Based on performance and FI measured from 0 to 70d, the average RFI was calculated and steers were divided into low- (n=14) and high-RFI (n= 15) groups based on 0.5 SD below and above the mean RFI, respectively. Untargeted serum metabolomics was conducted utilizing the Dionex UltiMate 3000 UPLC system and elec-trospray ionization was used to introduce the samples into an Exactive Plus Orbitrap MS. Genomic DNA was extracted from rumen content and the amplified VI-V3 hypervariable region of the bacterial 16S rRNA gene was sequenced for analyses. Missing values were approximated through matrix completion and data was normalized using a centered log-ratio transformation. Random Forests supervised machine learning and feature selection was performed on the bacterial compositions. Residual feed intake was associated with several attributes of the rumen bacteriome. Low-RFI steers were associated with decreased bacterial a- (P = 0.03) and p- diversity (R2 = 1, P = 0.001). Several serum metabolites were associated with RFI. Based on fold change (high/low RFI), low-RFI steers had greater abundances of pantothenate (0.375; P = 0.04) and reduced abundances of glucose-6-phosphate (2.13; P = 0.02) and glucose-1-phosphate (2.13; P = 0.03). Machine learning on RFI was highly predictive of both serum metabolomic signature and rumen bacterial composition (accuracy >0.7). Fold change Flavobacteriia abundances were greater with increased pantothenate contrasted to reduced pantothenate (5.06; P = 0.04). Greater abundances of pantothenate-producing bacteria, such as Flavobacteriia, may result in improved nutrient utilization in low-RFI steers. Pantothenate