Individual peanut seeds were studied concerning the feasibility of oil content evaluations by the Nuclear Magnetic Resonance method (NMR). Correlation and regression analyses with the extractive (Soxhlet) method were performed. A variance study among seeds per plant and seeds among plants was also made in order to access the possibility of selecting for oil content based on individual seed evaluations. The data showed a high and positive correlation (r = 0.98) between both methods. Although the values obtained by the Soxhlet method were slightly higher than those obtained by the NMR, the relationship between both methods was linear, indicating the possibility of correcting the data when necessary. The observed variances in oil content among seeds of the same plant were at least equal to those observed among plants. These values indicated that, for deviations in oil content of up to 2%, a number of seeds always higher than 1 is necessary for sampling.
peanut seeds; oil content; nuclear magnetic resonance