Poljoprivreda (Jun 2007)
NIRS PREDICTION FOR PROTEIN AND INTRAMUSCULAR FAT CONTENT OF RABBIT HIND LEG MEAT
Abstract
The goal of this study was to develop calibration equations to predict the chemical composition of raw, homogenized rabbit meat by means of near infrared spectroscopy (NIRS). 44 Pannon White rabbits were housed in groups in three different pen types (16 anim./m2), and were fed the same diet. Another 45 animals were housed in cages (12 anim./m2) and fed by different feeding regimes. Rabbits were slaughtered at the bodyweight of 2.4-2.5 kg. Homogenized fresh and freeze-dried left total hind leg muscles were investigated by NIRS using a NIRSystem 6500 equipment with small ring cup sample holder. The ether extract and protein content of all samples were determined chemically. Samples 44 of housing experiment were applied in producing LOCAL calibration equations tested on the 45 samples from the separate feeding experiment. Coefficients of determination (R2) of the predictions were 0.89 and 0.99 for fat, 0.85 and 0.96 for protein in fresh and freeze-dried samples, respectively. Results are reassuring, because the equations were applicable, however the analyzed samples were from independent housing and feeding systems. Therefore the chemical compositions differed in the two datasets, i.e. 9.46%, and 11.79% for fat, 85.75% and 83.44% for protein content in calibration and prediction datasets, respectively. The average of NIRS predicted values for fat and protein was 11.36%, 83.88% or 11.54%, 83.45% when using fresh or freeze-dried samples, respectively.
