Nondestructive phenotyping fatty acid trait of single soybean seeds using reflective hyperspectral imagery

Abstract

AbstractSoybean (Glycine max) is one of the most economically important crops in the world and is used widely for different purposes. Breeding programs have developed new varieties with desired traits, including altered fatty acid profiles and levitated protein content. The breeding process involves the selection of elite genotypes from a large number of experimental lines, which is time‐consuming and costly. This study aimed to evaluate the feasibility of a reflective hyperspectral imaging system in classifying the fatty acid content of single soybean seeds for high‐throughput phenotyping high oleic soybean cultivars. A hyperspectral imager was used to take hyperspectral images of 100 soybean seeds of two varieties. After comparing the performance of different methods for data preprocessing, featured spectral patterns were extracted to predict oleic acid and linoleic acid content using a successive projection algorithm. Results showed that the spectral reflectance of soybean seeds was able to differentiate different cultivars. The spectral information of selected regions of interest from the hyperspectral images of single seeds could reflect the spectral information of whole seeds. It is found that the classification models based on raw spectra performed the best in estimating the content of oleic acid and linoleic acid. The classification accuracy of the calibration for oleic acid and linoleic acid was both 100%, and the classification accuracy of validation was 90 and 93.3%, respectively. The results indicate that the reflective hyperspectral imaging technology might be used as a rapid and nondestructive tool to quantify the oleic acid content and linoleic acid content of single soybean seeds for high‐throughput phenotyping high oleic soybean.Practical ApplicationsIt is the interest of soybean breeders to develop high oleic soybean varieties that are in great demand. Studies indicate that the major factors affecting soybean oil content are oleic acid and linoleic acid. However, the nutritional values of soybean seeds, such as sugar, protein, oil, and fatty acids, are commonly measured using wet chemistry analysis, which is destructive, time‐consuming, and expensive. Although NIR spectroscopy can improve efficiency and simplify the measurement procedure, it still requires a large number of seeds. In breeding programs, a single seed with known oleic content is needed for further seed development. Bulk‐seed measurement using spectroscopy is not selective and would not determine the oleic profile of single seeds. Therefore, there is a pressing need to develop a high‐throughput and nondestructive method to quantify the nutritional values of individual seeds for precision breeding and improve the efficiency of breeding programs.