Prediction of nitrogen use in dairy cattle: a multivariate Bayesian approach

K.F. Reed, L.E. Moraes, J.G. Fadel, D.P. Casper, J. Dijkstra, J. France, E. Kebreab

Research output: Contribution to journalArticleAcademicpeer-review

5 Citations (Scopus)

Abstract

Quantification of dairy cattle nitrogen (N) excretion and secretion is necessary to improve the efficiency with which feed N is converted to milk N (ENU). Faecal and urinary N excretion and milk N secretion are correlated with each other and thus are more accurately described by a multivariate model that can accommodate the covariance between the three observations than by three separate univariate models. Further, by simultaneously predicting the three routes of excretion and taking advantage of the mass balance relationships between them, covariate effects on N partitioning from feed to faeces and absorbed N and from absorbed N to milk and urine N and animal ENU can be estimated. A database containing 1094 lactating dairy cow observations collated from indirect calorimetry experiments was used for model development. Dietary metabolisable energy content (ME, MJ/kg DM) increased ENU at a decreasing rate, increased the efficiency with which feed N was converted to absorbed N and decreased the efficiency with which absorbed N was converted to milk N. However, the parameter estimate of the effect of ME on post-absorption efficiency was not different from zero when the model was fitted to a data subset in which net energy and metabolisable protein were at or above requirement. This suggests the effect of ME on post-absorption N use is dependent on the energy status of the animal.
Original languageEnglish
Pages (from-to)1918-1926
JournalAnimal Production Science
Volume54
Issue number12
DOIs
Publication statusPublished - 2014

Keywords

  • cows
  • excretion
  • protein
  • management
  • metabolism
  • efficiency
  • ruminants
  • pollution
  • dietary
  • manure

Fingerprint Dive into the research topics of 'Prediction of nitrogen use in dairy cattle: a multivariate Bayesian approach'. Together they form a unique fingerprint.

Cite this