Predicting the dynamics of enteric methane emissions based on intake kinetic patterns in dairy cows fed diets containing either wheat or corn

Abstract

The production of methane by the rumen microbiota is a complex biological process. When tackling the modelling of methane production, the modeller decides what complexity is needed to answer the scientific question for which the model is intended. Such a choice results in a diversity of possible models spanning both empirical and mechanistic approaches. Within the framework of precision livestock farming, simple dynamic models offer great advantages for integrating online data (e.g., feed intake) to predict individual methane emissions from cattle. Accordingly, we previously developed, with satisfactory results, a simple dynamic model that uses DM intake kinetics as a single predictor of methane emissions from finishing beef steers. The objective of the present work was to assess the capability of the previously developed model to predict the dynamic pattern of methane production from dairy cows fed a diet containing either wheat grain or corn grain. We showed that the simple dynamic model in its original form enables a description of the dynamics of individual methane emissions from dairy cows with an average determination coefficient (r²) of 0.65 and an average concordance correlation coefficient of 0.81 and RMSE of 16% and 26% for the corn-based and wheat-based diets, respectively. Additionally, we performed a principal component analysis associating the parameters of the methane model with variables characterising the feeding behaviour of the cows. The results showed the effect of the diet type on the feeding behaviour of the animals. This impact was propagated on the dynamics of methane emissions. Interestingly, our model enabled us to determine that the differences in patterns of methane emissions between the diets result simply from the dependency of the methane yield and rate constant of methane eructation on the grain type.