One of the most relevant but expensive methods of assessing the quality of fish meal is the physiological digestibility test with minks. The purpose of this study was to determine whether spectroscopic and chemical analyses evaluated with chemometrics can replace minks in digestibility tests. The spectroscopic methods used were the two complementary techniques of fluorescence emission and near-infrared reflectance. The investigation included 54 samples of high-quality fish meal ranging from 89.6 to 93.9 on the mink digestibility index. The investigation also included determination of seven quality parameters in the fish meal to substantiate the spectroscopic models on the mink digestibility. These quality parameters include the content of protein, oil, water, water-soluble protein, ash and the biogenic substance cadaverine as well as the titration value. The study demonstrates that the mink digestibility could be predicted from combining the two reference quality parameters oil and ash with a correlation coefficient of 0.89. The best spectroscopic calibration models on the mink digestibility resulted in a correlation coefficient of 0.94 and a prediction error of 0.5 using the near-infrared spectral ensemble and a correlation coefficient of 0.92 and a prediction error of 0.5 using the fluorescence spectral ensemble. In addition, the seven reference quality parameters could be predicted from the spectroscopic ensembles with good precision. (C) 2000 Society of Chemical Industry.
|Journal||Journal of the Science of Food and Agriculture|
|Publication status||Published - 2000|