ABSTRACT

Sorghum is a critical source of food in the semiarid regions of sub-Sahara Africa and parts of India, and a potential source of dietary phytochemicals including carotenoids. A sorghum core collection representing a wide range of genetic diversity, and used in the framework of a sorghum breeding and genetic program was evaluated by near-infrared refl ectance spectroscopy to predict food grain quality traits: amylose content, protein content, lipid content, endosperm texture, and hardness (de Alencar Figueiredo et al., 2006). A total of 278 sorghum samples were scanned as whole and ground grain to develop calibration equations. Laboratory analyses were performed on near-infrared refl ectance spectroscopy sample subsets that presents the core collection racial distribution. Principal component analysis performed on these sample-spectra evidenced a level of structure following known sorghum races, which underlines the importance of using a wide range of genetic diversity. Performances of calibration equations were evaluated by the coeffi cient of determination, bias, standard error of laboratory, and ratio of performance deviation. Ground grain spectra gave better calibration equations than whole grain. Protein content equation (ratio of performance deviation of 5.7) can be used for quality control. Endosperm texture, lipid content, and hardness equations (ratio of performance deviation of 2.9, 2.6, and 2.6, respectively) can be used for screening steps. Even with a small standard error of laboratory in whole sample analysis, a ratio of performance deviation of 1.8 for amylose content confi rmed that this variable is not easy to predict with near-infrared refl ectance spectroscopy. The important foods prepared with sorghum are tortilla, porridge, couscous and baked goods.