| The rapeseed is one kind of important oil-bearing crops in China. Both its output and acreage remain the first place in the world. Rapeseed oil is one sort of important edible oil. Its meal contains rich protein and is used as the animal feeding stuff. But the glucosinolate content in traditional cultivars is relatively high, which has greatly limited its utilization in the feeding stuff. The genetic improvement is an effective way to reduce glucosinolate content of the rapeseed. The development and modification of analysis methods is closely connected with the qualities improvement. Therefore, the development of the simple and quick and precise quality analysis method has become a great subject that the breeders are concerned about.Being the quick and precise method to analyses complicated matter content developed in recent years, NIRS.which possesses superiorities such as non- destructiveness and safety etc, is a kind of green analysis technology.The purpose of this study was to determine the oil and glucosinolate content (including total glucosinolate and gluconapin, progoitrin, and 4- hydroxyglucobrassicin) by NIRS. The samples were analysed by the residual technique (oil content) and high performance liquid chromatography ( glucosinolate content ) as the reference methods. And the calibration models for the oil and glucosinolates were developed and evaluated respectively.Tibet ,where there are very rich germplasms which show many particular phonetypes,is one of the origin center of the B.campestris , and is also the probable origin center of the B.Juncea. To make clear the quality characteristics of these germplasms is important to enrich breeding resources and to effectively use their special characteristics. In this text, 529 Tibet rapeseed resources had been used to proceed breeding and to determine the content of oil, fatty acid and glucosinolate. The major conclusions of this research are as follows:1. NIRS calibration models for oil contentOil model 1 Oil model 1 contains 189 samples (harvested in 2001 and in 2002 ) ,the oil content of which ranges from 32.69% to 49.80%. Coefficient of determination (R2) is of 0.97, and the relative standard error of cross-validation (RMSECV) is of 0.66 %.0 i I mode I 2 Oil model 2 contains 144 samples (harvested in 2001), the oil content of which ranges from 30.42% to 49.34%. R2 is of 0.98 and RMSECV is of 0.60%. The predicted value by oil model 1 is precised than that by oil model 2 when forecasting unknown samples harvested in 2002.2. NIRS calibration models for total glucosinolate contentNIRS calibration models for samples with low and high total glucosinolate(model 1) . The model was developed from a calibration set of 43 samples (glucosinolate content: 14.67-174.55 μmol.g-1) with R2 of 0.99 and RMSECV of 4.66 μ mol.g-1NIRS calibration models for samples with low total glucosinolate, (model 2). The model was developed from a calibration set of 76 samples (glucosinolate content: 12.20-40.02μ mol.g-1) with R2 of 0.92 and RMSECV of 1.78μ mol.g-1.NIRS calibration models for samples with low and moderate total glucosinolate(model3) .The model was developed from a calibration set of 121 samples (glucosinolate content: 12.46-58.96 μmol. g-1) with R2 of 0.92 and RMSECV of 2.84 μmol .g '.3. NIRS calibration models for individual glucosinolates contentGluconapin, progonitrin and 4-hydroxyglucobrassicin are dominant in B.napus. The models for gluconapin, progonitrin and 4-hydroxyglucobrassicin were developed from calibration sets of 79, 81 and 81 samples with R2 of 0.92 ,0.95 and 0.87, and RMSECV of 1.11 μmol.g-1, 0.51μmol. g-1 and 0.47μmol.g-1, respectively.4. The contents of oi I , total glucosinolate and euric acid in Tibet B.campestris were 24.76%-45.51% , 30.57-192.2μ mol.g-1 and 21,03%-54.06% respectively; and that in B.juncea were 25.76%-43.03%, 37.59-145.80 u mol .gM and 27.11%-45.21% respectively. Among them Oil content of two materials in B.campestris was more than 45%;and also euric ac...
|
PDF Full Text Request