Evaluating near-infrared spectroscopy as a rapid and cost-effective method for fatty acid profiling in Senegalese sole (Solea senegalensis)

This study evaluates the potential of near-infrared spectroscopy (NIRS) as a rapid and cost-effective technique for estimating the fatty acid profile of Senegalese sole (Solea senegalensis). Traditionally, lipid profiling in aquaculture products has relied on gas chromatography (GC), a highly accurate but time-consuming, costly and complex method requiring extensive sample preparation. In contrast, NIRS offers a non-invasive, rapid and accurate alternative for predicting chemical composition. While NIRS has shown success in lipid profiling for fatty fish species, its application to lean fish, such as Senegalese sole, presents challenges due to low lipid concentrations and overlapping spectral signals. In this study, Senegalese sole were fed experimental diets varying in marine protein and lipid content. Lipid profiles of 75 tissue samples were analysed using both GC and NIRS, with fatty acid content expressed as percentage of total fatty acids (% TFA) and as µg fatty acid per mg dry tissue (µg FA/mg dry tissue). Calibrations were developed using Modified Partial Least Squares Regression (MPLSR) and the prediction accuracy was assessed using calibration statistics. The results showed that dietary lipids significantly influenced the fatty acid composition of Senegalese sole, consistent with previous findings for marine fish. NIRS calibrations were most accurate when fatty acid content was expressed as µg FA/mg dry tissue, giving better predictions than % TFA. The best calibrations were obtained for oleic acid, the dominant fatty acid in Senegalese sole, with an R²cv of 0.83 and a ratio of standard deviation/cross validation error (RPD) of 2.44. Fatty acid group predictions were more accurate for monounsaturated fatty acids (MUFA) than for the other groups, with an R²cv of 0.85 and an RPD of 2.60. These results demonstrate the potential of NIRS as a rapid and cost-effective method for fatty acid profiling in Senegalese sole, particularly for the estimation of mono- and polyunsaturated fatty acids (MUFA and PUFA). However, further research is needed to improve the accuracy and robustness of NIRS prediction models. Expanding the calibration dataset to include greater variability in diet, season and cultivation conditions will improve the reliability of the method for wider applications.