Effects of iso-alpha-acids on rumen fermentation and comparison of microbial populations between rumens and continuous culture fermenters

University of Minnesota M.S. thesis. August 2015. Major: Animal Sciences. Advisor: Marshall Stern. 1 computer file (PDF); x, 160 pages.Two experiments were conducted using dual-flow continuous culture fermenters to study rumen fermentation in vitro. The first experiment examined the effects of iso-alpha-acids from Humulus lupulus (Hops) extract on rumen fermentation. Iso-alpha-acid extract (IE) was provided to continuous culture fermenters to supply 0 (CON), 600 (LOW), 1200 (MED) or 1800 mg of IE/kg of diet DM/day. There was no effect (P > 0.05) of IE inclusion on DM, OM, NDF or ADF digestion (%). Volatile fatty acid (VFA) metabolism was not affected by IE treatment (P > 0.05), with total VFA concentrations of 105.5, 93.4, 87.9 and 103.6 mM for the CON, LOW, MED and HIGH treatments, respectively. Nitrogen metabolism was also not affected (P > 0.05) by IE level, with the CON, LOW MED, and HIGH treatments resulting in nitrogen concentrations of 7.4, 5.3, 7.6 and 6.8 mg N/dL of rumen fluid, respectively. No effects (P > 0.05) of treatment on fermenter pH were observed. It was concluded that administration of IE had no impact on fermentation by ruminal microbes maintained in continuous culture fermenters. Experiment 2 used 16S amplicon sequencing to compare microbial populations between the rumen of dairy cattle and continuous culture fermenters and to determine the temporal changes in microbial community during fermenter operation. Redundancy analysis (RDA) was performed to determine correlations between fermentation measurements based on microbial community. Correlations were also conducted to determine associations between prominent microbial families and fermentation measures from the In vitro system. Differences in microbial community were assessed using UniFrac metrics, Analysis of molecular variance (AMOVA) and analysis of similarity (ANOSIM) based on Bray-Curtis dissimilarity matrices. Differences in taxonomic composition of different sample types were analyzed at kingdom, phylum, class, order and family taxonomic levels. Functional inferences were made by matching taxonomic data to KEGG Orthology terms using PICRUSt software, and analyzed based on sample type. Community profile did not differ (P > 0.10) between cows in either rumen or inoculum samples, but was different (P 0.10) based on sample type. Proteobacteria, Tenericutes, Spirochaetes and Verrucomicrobia were different (P 0.10) in relative abundance of Prevotellaceae, which was the most abundant family in all three samples. Abundant families that were different (P 0.10) by sample type according to PICRUSt inferences