NIR reflectance spectroscopic analysis of the FA composition in sesame (Sesamum indicum L.) seeds

The feasibility of using NIR reflectance spectroscopy to estimate the FA composition of sesame seed (Sesamum indicum L.) samples from the National Institute of Crop Science of Japan and from Myanmar was examined. Multiple linear-regression analyses of NIR sepctral data and chemical data for whole seeds were carried out to develop calibration equations for predicting the proportion of each of the four major FA in sesame seeds from the total FA composition. The SE of prediction (SEP) was 0.616% for palmitic acid, 0.348% for stearic acid, 1.051% for oleic acid, and 0.826% for linoleic acid. This NIR method provides a simple, rapid, and nondestructive means of estimating the FA composition of sesame seeds for breeding selection, regardless of the color of the sesame seed coats. However, the proportions of palmitic and stearic acids could not be reliably measured because their SEP were almost as great as the SD of their concentrations in the set of prediction samples. The relationship between NIR spectral patterns and the FA composition of sesame seeds also was examined. The correlation coefficient calculated for the standardized second-derivative NIR spectral readings at 1708 nm and the percentages of linoleic acid was −0.830. A rough estimate of the proportion of linoleic acid in the total FA composition of sesame seeds could be obtained even with single sesame seeds, except for those with a black coat, based on NIR spectral pattern analysis using the wavelength assignments of linoleic acid.