PSII-B-15 Composition of Milk as Explained by Rumen Fermentation Parameters, Diet Composition and Combination

Abstract

Listen

Translate article

Abstract Although individual studies on the milk component responses to diet and rumen fermentation parameters, notably rumen pH, have identified relationships, these have not been widely summarized. Therefore, a meta-analysis was conducted with the objective of evaluating how milk fat and protein percentage change in response to diet (D), rumen fermentation parameters (R) and combination of both (R+D) variables. A secondary objective was to evaluate how these models differed when considering interactions with rumen pH. Rumen fermentation parameters included total volatile fatty acid (VFA) concentration, molar proportions of major and minor VFA, rumen pH and ammonia nitrogen, and diet variables included dry matter intake (DMI) dietary percentages of dry (DM) and organic matter (OM), crude protein (CP), neutral (NDF) and acid detergent fiber (ADF), ether extract, and starch. The database consisted of 138 published articles. Models were derived using a mixed linear regression weighted by the inverse of the standard error. Models were compared based on the concordance correlation coefficient after removing study effects (CCC). Models of milk fat based on R, R+D and R*pH+D*pH had similar capacity to explain variation within the data set (CCC: 0.26 to 0.31). Other than CP all other diet variables remained in the final D model and the same was observed in the model D*pH. Propionate was a significant variable in all R based final models along with acetate in the R*pH model and valerate in R model. Despite numerous significant relationships, ability to explain variability within the dataset was poor. All milk protein models had similar, and poor capacity to explain variability within the dataset (CCC > 0.2). Interestingly minor VFAs, valerate and iso-butyrate, were included in all R-based models. Although individual studies show precise dietary control of milk components, further work is needed to explain variability in those responses across studies.