Mixed Acid Fermentation of Carbohydrate-Rich Dairy Manure Hydrolysate

Abstract

Dairy manure (DM) is an abundant agricultural residue that is largely composed of lignocellulosic biomass. The aim of this study was to investigate if carbon derived from DM fibers can be recovered as medium-chain fatty acids (MCFAs), which are mixed culture fermentation products of economic interest. DM fibers were subjected to combinations of physical, enzymatic, chemical, and thermochemical pretreatments to evaluate the possibility of producing carbohydrate-rich hydrolysates suitable for microbial fermentation by mixed cultures. Among the pretreatments tested, decrystalization dilute acid pretreatment (DCDA) produced the highest concentrations of glucose and xylose, and was selected for further experiments. Bioreactors fed DCDA hydrolysate were operated. Acetic acid and butyric acid comprised the majority of end products during operation of the bioreactors. MCFAs were transiently produced at a maximum concentration of 0.17 mg COD_MCFAs/mg COD_Total. Analyses of the microbial communities in the bioreactors suggest that lactic acid bacteria, Megasphaera, and Caproiciproducens were involved in MCFA and C4 production during DCDA hydrolysate metabolism.

Keywords: biomass pretreatment; dairy manure; lignocellulosic biomass; medium-chain fatty acids; mixed culture fermentation.

FIGURE 1

Strategies for pretreatment of the lignocellulosic fraction of DM used in this study. Scheme 1 consisted of seven total unit operations and included physical, chemical, and enzymatic steps. Scheme 2 consisted of five total unit operations and included physical and thermochemical steps. The grey arrows and text indicate streams of material that were neither used in downstream bioprocessing units nor analyzed. Abbreviations: liquids-solids (L/S).

FIGURE 2

Composition of DM fibers on a weight basis. Percentages shown are averages from both Scheme 1 and Scheme 2 batches of DM.

FIGURE 3

Batch fermentation of DCDA hydrolysate shows simultaneous glucose and xylose consumption with transient accumulation of L-lactic acid.

FIGURE 4

Fermentation experiments with DCDA hydrolysate. One bioreactor experiment (DMB) was ran for 42 days with sampling every 6 days–panels (A) and (C). A second bioreactor experiment (DMB2) was ran for 10 days with frequent sampling–panels (B) and (D). The concentrations of quantified extracellular metabolites are presented in panels (A) and (B) for DMB and DMB2, respectively. Bars on Day 0 of panels (A) and (B) are the measured concentrations at start up and reflect the concentrations in the sludge inoculum. Dashed lines reflect the strength of the influent (black) and inoculum (red) in terms of their COD. The fraction of initial glucose, xylose, and other soluble carbohydrates that was utilized in DMB and DMB2 are presented in panels (C) and (D), respectively. Residual soluble carbohydrates accounts for the total concentration of unconsumed soluble carbohydrates including glucose and xylose. Concentrations of other soluble carbohydrates were calculated by subtracting concentrations of glucose and xylose from total soluble carbohydrates. Soluble uncharacterized COD was calculated as the soluble COD subtracted by the combined COD of all measured metabolites. Insoluble uncharacterized COD was calculated as the total uncharacterized COD subtracted by soluble uncharacterized COD. Error bars in all plots reflect the standard deviation of analytical replicates.

FIGURE 5

Phylogenetic tree and heatmap of ASVs with relative abundances >1% in the inoculum or at one or more timepoints during DMB and DMB2 operations. Branch labels indicate bootstrap values; values <50 are not shown. The scale bar indicates the branch length (solid-lined) at which 0.5 changes per nucleotide is estimated. Assignments ending in “sp.” indicate an assignment to an uncultured bacterium. Assignments ending in “unclassified” denote ASVs that were unable to be classified at the species-level, and in one case (CLOS101) at the genus-level. Parenthetical numeric values indicate the confidence value of the classification of the species, genus if the species-level is not given, or family if the genus-level is not given. Parenthetical abbreviations are manually assigned monikers. A bar plot of the sum of abundances of displayed assignments is shown atop the heatmap. The phylum (bold) and class (colored boxes) of ASVs are shown right of the heatmap. Abbreviations: Firmicutes (F), Epsilonbacteraeota (E), Proteobacteria (P), Actinobacteria (A).

FIGURE 6

RDA biplot illustrating microbial and metabolomic data of DMB and DMB2. Circles represent sample points that were taken during reactor operation (orange: DMB; green: DMB2), numbers adjacent to circles indicate the sampling day to which the circles correspond, crosses to ASVs (ASVs that are ordinated away from the center are labeled by their monikers in red; Supplementary Table S4), and vectors to explanatory variables. The percentage of total variation that each axis represents is indicated in parentheses within the axis titles.