Peanut(Arachis hypogaea L.),as an important crop,has rich nutrition in seed kernels and many ac-tive substances in seed coat.It is considered that kernels with dark seed coat are benefit to human health for its high-er anthocyanin content.A simple and easy prediction method on the anthocyanin content is imperative for breeding new varieties with high anthocyanin content.In this study,the F2 population derived from crossing with peanut varie-ty Yuhua 91(pink seed coat with anthocyanin content 182.33 mg/g)and Jihuatian No.1(purple seed coat with an-thocyanin content 270.69 mg/g)was used as the modeling material.The spectrally data of 194 peanut kernels were collected by the Antaris Ⅱ Fourier transform near-infrared spectroscopy,and then the anthocyanin content of peanut seed coat was determined.The partial least squares(PLS)method was used to construct a near-infrared spectrosco-py model for rapid and non-destructive detection on anthocyanin content.The internal validation means square error of cross validation(RMSECV)of the model was 10.3,and the correlation coefficient R2=0.990 3.Twenty-four other peanut materials not involved in the modeling were selected for external verification of the model,and the coeffi-cient of determination of the model prediction value and chemical determination value R2=0.978 1,which indicated that the model could be applied to the rapid detection of anthocyanin content in peanut kernels.The model was used to screen the offsprings from population of Yuhua 91 and Jihuatian No.1,and 14 germplasms with high oleic acid and high anthocyanin content,16 germplasms with high sugar and high anthocyanin content were obtained suc-cessfully.
NETL
NSTL
万方数据
