| Yellow-seeded rapeseed(Brassica napus L.) is characterized with thinner seed coat and lower hull percentage,which in turn is correlated with higher oil,protein and lower crude fiber contents compared with black-seeded or brown-seeded rapeseed under the same genetic background.The breeding of yellow-seeded rapeseed has become one of the main objectives in rapeseed breeding.Although the genes controlling yellow-seed of B.napus are mainly derived from A genome of B.campestris,only a few from the C genome of B.olerasea.So,it is hard for breeders to select effectively yellow rapeseed lines due to the scarcity of yellow-seeded gene resources.It is very important to explore new genetic resources of yellow rapeseed gene.B.oleracea var.aceapala is,a variety of B.oleracea,Cruciferae,a two-year herbaceous flower.The Chongqing Engineering Research Center for Rapeseed,in Southwest University found three yellow-seed plants in the B.oleracea var.aceapala in 1994.They would be valuable genetic materials for breeding of yellow-seeded in B.napus.However,B.oleracea var.aceapala is always used as edible vegetable and ornamental plant and no research involved in its seed quality has been found up to now.In this study,the materials were offered by Chongqing Engineering Research Center for Rapeseed,spectroscopy data for oil,protein and fatty acid compositions in seeds B.oleracea var.acephala were collected by near infrared reflectance spectroscopy(NIRS).The contents of seed oil and meal protein of 250 lines and seven fatty acid compositions of 282 lines were determined by chemical methods of national standard for NIRS modeling.Predictive models of seed oil content and meal protein content and seven fatty acid compositions were established by using two regression methods of Partial Least Square(PLS) and Modified Partial Least Squares(MPLS) and different technologies of wavelengths and mathematic treatments.The aim is not only to enlarge the area of study and application of B.oleracea var. aceapala,but also to establish some foundation in utilization of B.oleracea var.aceapala in rapeseed breeding program.The main conclusions of this study had been as follows:1.Seed oil contents of the lines ranged from 20.52%to 40.05%with an average 29.55% in this study and almost all the lines were scattered between 24.00%and 36.00%.In the predictive model of seed oil content,internal cross validation test showed that 1 minus the variance ratio(1-VR) was 0.9180 and standard error of cross-validation(SECV) was 1.0606. In external test of the models,determination coefficient of validation(RSQ) was 0.949 and standard error of performance(SEP) was 0.881.So,the predictive models had high precision and can be used in quality breeding programs.2.Meal protein content of the lines varied from 36.39%to 50.96%and the mean value was 45.21%.There were about 91%of the lines distributing between 24.00%and 36.00%. The predictive model of meal protein content was initially evaluated by internal cross validation test showing a 1-VR of 0.9060 and a SECV of 0.8720.It was further evaluated using external test revealing a RSQ of 0.915 and a SEP of 0.7790.This indicated that the meal protein content predictive model had also high precision and could be widely used in quality breeding programs.3.Among the NIRS models of fatty-acid composition content,oleic acid and linoleic acid ware the best,their 1-VR and RSQ were more than 0.8;linolenic acid,eicosenoic acid and erucic acid were better,their 1-VR and RSQ ranged from 0.70 to 0.82;palmitic acid and stearic acid were bad,1-VR and RSQ were low 0.70.According to the above results it known that,the NIRS models of oleic acid and linoleic acid are better,but other fatty-acid models were not ideal value,these models should be improved.
|
PDF Full Text Request