PSV-14 Effects of Forage or Coproduct-Based Transition Diets on the Rumen Microbiome and Ruminal Fermentation ex Vivo in Feedlot Steers

Abstract The objective was to determine the effect of replacing forage with coproducts in transition diets on rumen microbiome composition and ruminal fermentation ex vivo after steers consumed a common finishing diet. A random subset of 30 steers were selected from a larger, separate pen study. Steers were blocked by initial BW and allotted into 9 pens. Pens were then assigned to 1 of 3 dietary treatments for the 35 d adaptation period: a forage-based diet (FO; alfalfa and grass hay), a coproduct-based diet (CO; soybean hulls and modified wet distillers grains with solubles), and a coproduct-based diet pair-fed (CO-P) to isoenergetic levels of FO. Dry-rolled corn inclusion increased with either decreasing forage or coproduct inclusion. Both FO and CO were fed with slick bunk management. At d 35 and 63, rumen fluid was collected by esophageal tubing to determine differences in fermentation of the finishing diet as the substrate in an ex vivo model. Rumen fluid was also utilized for microbiome analysis using 16S rRNA sequencing on the Illumina MiSeq platform with 2 × 300 bp paired-end sequencing. Sequencing reads were analyzed within QIIME2. Statistical analyses were conducted using SAS 9.4. Fermentation data were analyzed using the MIXED procedure, while bacterial relative abundance data were analyzed in the GLIMMIX procedure. Correlations were analyzed using the CORR procedure. A treatment × sampling day interaction (P = 0.02) was observed for in vitro dry matter disappearance; on d 35, it was greater in steers fed CO compared with CO-P (58 vs. 53%, respectively), but no differences were observed on d 63. Within the rumen microbiome, there was a tendency (P = 0.06) for alpha diversity to have a treatment × day interaction. Specifically, CO steers had the greatest decrease in alpha diversity from d 35 to 63, but there was no change in alpha diversity in FO steers and CO-P steers over time. Final pH in from the ex vivo model was positively correlated (r = 0.32) with Bacteroidetes relative abundance but negatively correlated (r = -0.25) with Proteobacteria (P ≤ 0.05). In vitro dry matter disappearance was positively correlated (r = 0.38) with relative abundance of Firmicutes, but negatively correlated (r = -0.43) with Proteobacteria (P < 0.01). Dietary treatment did not affect (P ≥ 0.21) richness within the rumen microbiome. Rumen microbiome composition and ex vivo fermentation results suggest all treatments were adapted for the finishing diet at the conclus