p.466
p.472
p.479
p.485
p.490
p.497
p.501
p.506
p.510
Predicting Main Fatty Acids, Oil and Protein Content in Intact Single Seeds of Groundnut by near Infrared Reflectance Spectroscopy
Abstract:
Eighty-four, 61 and 105 groundnut seeds, including high oleate genotypes and F2 seeds of normal oleate × high oleate crosses, genotypes with high or low oil and F2 seeds from high oil × low oil crosses, and randomly selected samples representing various origins, different seed sizes, and varied seed coat color and protein content, were used to develop the NIRS models for main fatty acids, oil and protein. For oleic, linoleic and palmitic acid, the optimized spectrum pretreatment method was first derivative plus multiplicative scattering correction; for stearic acid and the four bad fatty acids, first derivative plus vector normalization. The Rcal2 and RMSECV for oleic acid were 97.20% and 2.65%; for linoleic acid, 96.90% and 2.40%; and for palmitic acid, 93.39% and 0.53%, respectively. The best spectrum pretreatment method for oil and protein was first derivative plus multiplicative scattering correction and min-max normalization. For oil and protein, the Rcal2 was 89.06% and 91.45%, and RMSECV, 0.89% and 0.78%, respectively. The NIRS models can be used to develop groundnut cultivars both with high oil and with high oleate to cater the growing need for biodiesel production.
Info:
Periodical:
Pages:
490-496
Citation:
Online since:
December 2013
Authors:
Keywords:
Price:
Сopyright:
© 2014 Trans Tech Publications Ltd. All Rights Reserved
Share:
Citation: