Non-destructive method for selection of soybean lines contained high protein and oil by near infrared reflectance spectroscopy.

摘要

The applicability of the non-destructive near infrared reflectance spectroscopic (NIRS) method in determining the protein and oil contents of intact soyabean seeds was investigated. Seed samples derived from 310 soyabean germplasm lines grown in Korea during 2000 were measured as intact seeds by the NIRSystem Model 6500 monochrometer near infrared reflectance spectrophotometer using a small reflectance vessel. For protein content, the difference of scatter correction method, wavelength and math treatment effect did not highly improve the MPLS regression performance. Thus, the optimum condition was obtained at 2, 8, 6, 1 (2nd derivative, 8 nm gap, 6 points smoothing and 1 point second smoothing) math treatment condition with standard normal variateand detrend (SNVD) scatter correction method and entire spectrum. For the oil content, the best equation condition was obtained at 2, 1, 20, 10 (2nd derivative, 1 nm gap, 20 points smoothing and 10 points second smoothing) math treatment condition withSNVD scatter correction method and entire spectrum. The validation of these nondestructive NIRS equations showed very low bias (protein: 0.060%, oil: 0.451%) and standard error of prediction (protein: 0.568%, oil: 0.451%) as well as high coefficient of determination (protein: 0.927, oil: 0.906). These suggest the reliability of the NIRS equations in determining the protein and oil contents of intact seeds, and the applicability of the NIRS method as a mass screening technique for the selection of soyabean with high protein and oil content.