Comparison of energy evaluation systems for dairy cow feeds

• Published in: Livestock production science 1997-11, Vol.51 (1), p.255-266
• Main Authors: Kaustell, K., Tuori, M., Huhtanen, P.
• Format: Article
• Language: English
• Subjects: Dairy cows; Energy; Feed evaluation
• ISSN: 0301-6226; 1872-6070
• DOI: 10.1016/S0301-6226(97)00054-7

Nine energy evaluation systems were compared on the basis of results from 51 milk production trials and 261 diets. The nine systems evaluated were the fattening feed unit used in Finland (FU F), the feed unit used in Denmark (FU D), the metabolizable energies used in the UK (ME MAFF, ME ARC) and Sweden (ME SW), the net energy in lactation used in the Netherlands and Norway (NEL NL), France (NEL FR) and the USA (NEL US) and the net energy in fattening from Rostock (NEF RO). Energy values for feedstuffs were calculated using the same digestibility coefficients in each system. Energy supply ( S), requirements ( R) and the S/ R ratio were calculated using the treatment means. The accuracy of the systems was studied through the one-way analysis of variance with the trial as a classifier. Small residual variance (MSE) in the S/ R ratio indicated that a system accurately predicted the differences in the productive values of the diets within the trial. The relation between estimated and observed milk yield was studied through regression analysis. The MSE of the S/ R ratio was smallest in ME MAFF and NEL NL. The regression coefficient between the estimated and observed milk yields was closest to one and the intercept closest to zero in ME MAFF and in NEL NL. It is concluded that with Finnish dairy cow diets, ME MAFF is the most accurate of the feed evaluation systems examined; the next best are NEL NL and NEL US.