Spectral signature for quality assessment of anchovy fish meal (Engraulis Ringens) using Partial Least Squares

Abstract

Fishmeal is one of the most important foods for the aquaculture, livestock and poultry sectors in Latin America and the world. However, one of the constraints in the fishing industry is to measure online the physicochemical composition of this product and ensure its quality. In this study, the spectral signature of anchovy fish meal (Engraulis Ringens) has been obtained in wavelengths from 400 to 900 nm and has been related to its values of Protein, Ash, Moisture, Fat, Free Fatty Acids (FFA), total volatile basic nitrogen (TVB-N), sand, histamine and antioxidant remnant (REM.A/O). To analyze these relationships has been used the regression of Partial Least Squares (PLS) finding the best results to determine the Protein, Fat, Moisture and with some accuracy TBVN and Histamine. In addition, the most important spectral ranges have been identified for the calculation of the estimates. With these results we have a model to determine the physicochemical composition of anchovy fish meal based on the spectral signature that could be implemented in automatic systems for process control.