Measurement of Metabolizable Energy, Gross Energy, and Moisture in Feed Grade Fats by Near Infrared Reflectance Spectroscopy

Abstract Near infrared reflectance analysis (NIRA) was assessed as a potential rapid measure of AMEn, gross energy (GE), and moisture content of feed grade fats. Values for AMEn were obtained for 41 fats and fat blends using a total collection procedure that included four levels of fat inclusion: 3, 6, 9, and 12%, respectively. Three oils (corn oil, soybean oil, and canola oil) showed AMEn values that approximated their GE. The AMEn, GE, and moisture values ranged from 5.961 to 9.664 kcal/g, 8.623 to 9.831 kcal/g, and .001 to 1.63%, respectively. A scanner NIRA spectrophotometer was used to obtain the spectra of the fats. The best algorithm (mathematical treatment of spectra) for the NIRA determination of AMEn and moisture was the second derivative (D2 OD), whereas GE was best measured after spectra was transformed into first derivative (D1 OD). Accuracy of the NIRA predictions was determined by the standard error of the estimate (SEE). The SEE values for AMEn, GE, and moisture for the validation set (samples not used in developing calibrations), were: .31 kcal/g, .08 kcal/g, and .12%, respectively. The NIRA technique showed the potential as a fast, nondestructive alternative method to measure AMEn and GE of fats and oils; however, NIRA was not good for predicting moisture in fats.