Application of near infrared reflectance (NIR) spectroscopy to predict carotenoid contents in sweet corn germplasm

Abstract

Adopting a fast and efficient near infrared spectroscopy (NIRs) for assessing carotenoid content in sweet corn germplasm would benefit to the breeding program. This study aimed to evaluate lutein, zeaxanthin, β-cryptoxanthin, β-carotene, and total carotenoid content in sweet corn germplasm by using the NIR technique coupled with multivariate analysis of partial least squares (PLS) regression and high performance liquid chromatography for constructing and validating the calibration models. The result indicated that NIRs can be used for screening and predicting carotenoid content in corn germplasm. There were high coefficients of determination (R² = 0.60-0.84) and a high ratio of standard error of prediction to standard deviation (2.14-2.53). While, the values for standard error of calibration, standard error of prediction, and bias were low. Furthermore, the carotenoid contents predicted by the developed calibration models were not significantly different from the reference values. Therefore, NIRs could be an alternative tool for evaluating carotenoid content in sweet corn germplasm for breeding programs.