Establishment of Calibration Models of NIRS Analysis for Cottonseed Oil and Protein Contents and Mapping Related QTLs in Upland Cotton

This study was primarily conducted to investigate the feasibility of near-infrared reflectance spectroscopy (NIRS) in the determination of cottonseed oil and protein contents. A total of 112 cotton lines were used to construct the calibration models of cottonseed oil and protein contents determined by NIRS, and the coefficients of determination R² of these two models were 0.9228 and 0.9249, respectively. The results showed that the NIRS methods tested in this study have relatively high accuracy in prediction test for cottonseed oil and protein contents. Two F₂ populations from the crosses DPLSR3 × Yancheng115 (Pop A) and Humcopedixi14wtr5 × Yancheng90-110 (Pop B) were employed as experimental materials to identify QTLs associated with cottonseed oil and protein contents and 4 associated QTL were found. A stable QTL for cottonseed oil content was mapped between the SSR maker DC20120 and NAU2893 in both F₂ and F₃ generations in PopA, explaining 9.21% and 12.01% of the phenotypic variation, respectively.