Influence of feed source on determination of fat and protein in milk by near-infrared spectroscopy

Abstract

,\ bstract R~.;el v~d 9 September 1998: accepted 10 April 1999 \ !v[lIk s ample~ fr om two experiments (II = 506 in tOlal) were an ,dyzed by measurement of near-infrared (NIR) spectra to investigate the etlect of feed source on milk fat and prolein content. ivlilk samples from the lirst experiment (n = 3uO) were used for development of the calibration equation and validation of the equation. The calibratlon equatIOns thus developed in the first experiment were then used in the second experiment to ~xamine the applications of NIR measurement. All feeding trials used the same basal rations, which conSisted of corn silage, [talian ryegrass, Alfalfa haycube, corn Rake and commercial concentrate. The first experiment was comprised of three rations: (I) basal ration , (2) basal ration with soy bean meal (48% of total crude protein (CP)) and (3) basal ration with soybean meal (19% of total CP) and fish meal (25% of total CP) The second experiment was comprised of five supplement alternatives (I) no supplement (NS), (2) corn g~uten meal (CGM, 26% of total CP), (3) nsh meal (FM, 26% of total CP), (4) defatted soybean meal (SSM , 28% of total CP\ and (5) roasted soybean meal (RSBM, 26% of total CPl. Feeding regimes in both experiments were adjusted to fulfill the maintenance and production requirements. The results showed that NIR prediction of milk fat content was not influenced by the feed of animal$. while the accuracy of protein prediction was significantly affected by the kind orreedstuff used in the ration . Thus, a wide range of milk sample~ from cows 0 11 ~ Iearl y defined feeding regimes is necessary for developing a satisfactory calibration for NIR systems. © [999 Elsevier Science Ltd . Ali rights reserved. Keywords .· NIP. analysis; Milk fal; Milk pro tein : }·eed supplemenl ation