400 Relationship between fecal microbiota and enteral methane emissions of dairy cows

Abstract

Abstract Relationships between enteral methane emissions and abundances of bacterial taxa in the feces of dairy cows were determined to assess if these abundances can be used to predict these emissions. Six mature non-lactating Holstein dairy cows on diets with forage (alfalfa/grass hay) to grain ratios of 100:0, 75:25, and 50:50 were used during 5-wk experimental periods in a replicated 3x3 Latin Square Design. Dietary NDF and starch concentrations ranged from 38.7 to 56.0 % DM and from 0.5 to 19.5 % DM among diets, respectively. Methane outputs were measured using an open-hood calorimetric system during two 24 h periods on two separate days during the fifth week of experimental periods. Daily methane emissions ranged from 288.5 to 588.5 L/d among cows and diets, and averaged 413.4 L/d. Feces were sampled twice daily on the days preceding methane measurements. Compositions of the microbiota in feces were determined using Illumina 16S rRNA sequencing. Linear regression models were developed using the MIXED procedure of the SAS to determine the relationships between daily methane emissions (L/d) and the abundances of bacterial taxa in feces (%). Relative abundances of taxa that were significantly correlated with methane emissions and that were present in at least 30% of feces samples were included in the initial model. Abundances with a significance level greater than 0.25 were stepwise removed from the model. Taxa with high correlation coefficients (r > 0.75) were not placed together in models. The final model included the abundances of 6 bacterial families in the feces, and had an R2, CV, and root MSE values of 0.92, 4.93, and 20.2, respectively. These results suggest that enteral methane emissions can be predicted from the bacterial composition of the feces. However, model validation with different data sets and diets is needed.