The Application Of Near Infrared Spectroscopy In Rapid Identification Of The Fish Meal Quality

Fish meal is a common and important protein feed ingredients, in which the amino acid ratio is acceptable, balanced, the amount of amino acids is abundant. Thus, it's commonly used in the feed production and processing. The adulteration of fish meal often occurs because of its good quality, nutritional value, largely used, high price.High-protein ingredients, conventional sensory, physical and chemical test results (such as crude protein, crude fat, moisture, ash, etc.) cannot describe the quality of fish meal yet, unless various amino acids content are shown.How to ensure the quality of feed and raw materials, improve the detection rate have become a recent and important topics to be discussed. Liquid Chromatography Automatic Online Pre-column Derivatization, Liquid Post-column Derivatization, Near-infrared Diffuse Reflectance Technique were used to determine 17 kinds of amino acid content in 145 fish meal samples. The results showed that:1. The use of new Automated Online Pre-column Derivatization method was more rapid, accurate, sensitive, and reproducible than the classic GB analysis method. Instrumental Analysis Experiment process was shortened to 25 minutes, and the fastest one can be shortened to 14 minutes. Besides, the prepare of mobile phase was simpler, only pH adjustment of buffer solution were needed. The derivation process can be implemented automatically.2. The precision and accuracy of post-column derivatization were high and separation was reliable, which characterized by simultaneous determination of 17 amino acids, and a smaller injury to amino acid analysis column. It could fully meet all the requirements of qualitative and quantitative analysis.3. The improved partial least squares (PLS) method was used in the Near Infrared Spectroscopy(NIRS) to build a NIRS calibration model of 17 amino acids in fish meal. The chemical analysis values and the determination coefficient R2 of NIRS predicted values of methionine, lysine, leucine, phenylalanine, aspartic acid, threonine, glutamic acid, glycine, alanine, arginine, proline, valine, isoleucine were above 0.9 and the relative standard deviation were less than 10%. The results showed that NIRS can be used to detect amino acids in fish meal rapidly and accurately.4. Based on the above results, the determination of amino acids in the feed and feed ingredients in our Laboratory are improved as following methods:I. The software of NIR3700 NIRS produced by FOSS company was upgraded, the calibration equation has been added to the module to achieve the rapid detection of amino acids in the fish meal; II. "Automatic On-line Pre-column Derivativatiom to determine amino acid" and " Post-column Derivativation to determine amino acid" were compiled as a method to determine two amino acids deviated from the standard.