Use of NMR for predicting protein concentration in soybean seeds based on oil measurements

Abstract

AbstractIncreasing the level of protein in soybean seeds has been a major target for soybean [Glycine max (L.) Merr.] breeders. The objective of this study was to examine the potential of predicting soybean seed protein based on oil values as determined by NMR. Seed protein and oil concentrations were determined in an F2 population generated from the cross between a G. max (NK, S08‐80), and a G. soja (PI 458536) cultivar. The protein concentration in the population ranged from 40.4 to 52.6%. Protein‐oil regression analysis was used to generate an equation for predicting seed protein concentration based on oil readings. The regression equation Protein=62.3–1.3 [Oil] (R2=0.46) was developed, with a corresponding correlation of −0.69 between the traits. With this equation, the mean protein concentration of the selected 25% of the population (a simulated breeding pressure) was greater than the mean of the unselected population (46.1%, SE=0.13) by about 1.9%. Individual F2 plants that exceeded the mean protein value of the population constituted 86.4% of the selected samples. Selection based on oil concentration, however, failed to include 27.1% of the plants that were among the top 25% for protein concentration. Selection of high‐protein plants based on NMR oil measurement was reasonably effective in the test population and might offer a new and rapid method of selecting high‐protein individuals in soybean populations derived from the wild soybean progenitor, G. soja. If further tested on other populations and samples, it might be used as an analytical alternative for an indirect measurement of protein concentration based on NMR measurements of the oil.