| Brassica oilseed crop is an important human and animal food source because of its high quality oil and high protein feed meal. The purpose of present study was to test the accuracy of near-infrared spectroscopy (NIRS) for estimating amino acids contents in rapeseed meal(Brassica napus L.). NIRSystems model 5000 monochromator and WinISI II software were applied to collect spectra and develop the calibration equations in this research. After the original population with 621 samples was scanned, two hundred and twenty-six samples of them were selected based on the spectra variation for the development of a calibration equations for different amino acid contents. Fifty-five samples were emploied as external validation set. The calibration equations were developed by using a standard normal variate+de-trending scatter correction and a 2,4,4,1 mathematical treatment as well as a modified partial least square (MPLS) regression method. Twelve amino acids (aspartic acid, threonine, serine, glutamic acid, glycine, isoleucine, leucine, tyrosine, phenylalanine, lysine, histidine and arginine) in rapeseed meal could be predicted with perfect precision (Coefficient of determination in calibration (RSQ)=0.877～0.976,1 minus the ratio of unexplained variance to total (1-VR)=0.861～0.971; coefficient of determination in external validation (R2)=0.840～0.953; SD/SECV (the ratio of SD (standard deviation) to SECV (standard error in cross-validation))=2.678～5.901; and SD/SEP (the ratio of SD to SEP (standard error in performance))=2.490～5.532). The equations for proline and methionine with moderate determination coefficients (RSQ=0.812,0.800; 1-VR=0.782,0.784;R2=0.814,0.706; SD/SECV=2.142,2.154; SD/SEP=2.311,1.851, respectively) showed better agreement between reference value and spectra by NIRS. The calibration model for alanine and cysteine (RSQ=0.839, 0.749; 1-VR=0.821,0.728; SD/SECV=2.365,1.921; R2=0.623,0.573; SD/SEP=1.633,1.521, respectively) could be used to predict the two amino acids content in breeding programs. The calibration model for valine (RSQ=0.620; 1-VR=0.579; SD/SECV=1.537; SD/SEP=1.243) however could explain less variation, and predicted the amino acid lacking of enough accuracy.For the developmental genetic analysis of rapeseed meal quality traits, a genetic model for quantitative traits of seed in diploid plant and corresponding unconditional and conditional statistical methods were applied for analyzing the genetic main effects and the genotype×environment (GE) interaction effects for rapeseed quality traits including lysine content, leucine content, isoleucine content, phenylalanine content, threonine content, glutamic acid content and glycine content. Nine parental lines of B. napus including Youcai 601, Double 20-4, Huashuang 3, Gaoyou 605, Zhongyou 821, Eyouchangjia, Zhong R-888, Tower and Zheshang 72 were used in a complete diallel design. Analysis of diploid embryo, cytoplasm and diploid maternal genetic effects and their GE interaction effects for rapeseed meal quality traits at different times/stages after anthesis were conducted by using two-year data. The genetic correlations among developmental times or among rapeseed quality traits at maturity were analyzed by using a mixed model approaches. The prediction for value of parents and the estimated for the heterosis of meal quality traits were also analyzed in present experiment.The results of unconditional variance analysis revealed that the amino acids quality traits of rapeseed meal were simultaneously affected by genetic main effects and their GE interaction effects from diploid embryo nuclear genes, cytoplasm genes and diploid maternal plant nuclear genes at different developmental times, meanwhile the results from conditional variance analysis showed that the new expression of genes from different genetic systems existed at different developmental stages. Among them, lysine content at 15d,22d,29d,36d or 43d after anthesis, leucine content at 15d,22d or 29d, isoleucine content at 15d,22d,29d or 36d, phenylalanine, threonine and glutamic acid contents at 15d,22d or 29d, glycine content at 15d,22d,29d or 43d were mainly controlled by GE interaction effects, while the performance at other developmental times for these amino acid traits were mainly governed by genetic main effects. Genetic analysis from the different genetic systems indicated that lysine, leucine, isoleucine, phenylalanine, threonine and glutamic acid contents were mainly controlled by maternal effects, followed.by embryo effects. Glycine content at 15d,22d,29d or 43d was mainly controlled by maternal effects but by embryo effects at 36d. The conditional variance analysis revealed that the new expression of genes especially at 23-29d after anthesis for lysine, threonine, glutamic acid and glycine contents, 1～15d for leucine content, and 16-22d for isoleucine and phenylalanine contents were relevant as the highest genetic activity were found during these developmental stages. Furthermore, net GE interaction effects from conditional analysis were more important for most nutrition quality traits at most developmental stages indicating new expression of genes was easily influenced by the environment conditions.Among the different genetic effects, additive effects were more important for most amino acids traits at most development times/stages. It was suggested based on the larger additive and cytoplasmic variances and their GE interaction variance that additive and cytoplasmic effects were more important than other genetic effects for lysine, isoleucine, phenylalanine, threonine, glutamic acid and glycine contents of rapeseed meal. Their total narrow-sense heritabilities were high, better effects for improving these nutrition quality traits could be achieved by selection in early generations. Otherwise, heritability analysis showed that the general heritabilities from different genetic systems were more important than interaction heritabilities for lysine at 36d after anthesis, and leucine, isoleucine, phenylalanine, threonine and glutamic acid contents at 36d and 43 d, thus selection for these nutrition traits was efficient and least likely to be affected by the environment. However, it was in the contrary at other development times for these amino acids traits in rapeseed, at all development times for glycine content.Heterosis including the general heterosis and the GE interaction heterosis from embryo, cytoplasm and maternal effects could be observed in most amino acids traits of rapeseed meal. The most important general heterosis for lysine content was embryo and maternal heterosis, while glutamic acid content was mainly affected by embryo heterosis. Furthermore, negative maternal heterosis was predominant of general heterosis for leucine, phenylalanine, isoleucine, threonine and glycine contents. The positive GE interaction heterosis was stable for isoleucine content across environment; it was not case for other amino acids traits because of the positive or negative GE interaction heterosis at different environment. Correlation analysis among different developmental times for rapeseed quality traits showed that the correlations tend to become more obvious, especially between 29 and 36d,29 and 43d,36 and 43d after flowering as rapeseeds develop. Significant relationships existed between different developmental times. Among them, phenotypic correlation (rP) and genotypic correlation (rG) between 15 and 29d for isoleucine, phenylalanine and glycine content,36 and 43d for lysine,29 and 36d,29 and 43d or 36 and 43d for leucine, isoleucine, phenylalanine, threonine, glutamic acid and glycine contents were significant positive. The GE interaction relationships from different genetic systems were more vital than genetic main relationships among the developmental times, and the relationships were easily influenced by the environment conditions. It was, in general, suggested based on the components of genetic main correlation (rGM) and GE interaction correlation (rGE) that embryo dominance main effect correlation (rD) and GE effect correlation (rDE), maternal dominance main effect correlation (rDm) and GE interaction correlation (rDmE) were more important than the others for the relationships of pairwise developmental times among different genetic systems across environments.It was suggested that the relationship could affect the indirect selection of multiple traits in a rapeseed breeding program by correlation analysis between different quality traits in mature rapeseed. The results showed that significant positive rP and rG were found between lysine and methionine/leucine/isoleucine/phenylalanine/threonine contents, indicating that genetic improvement in these 5 pairwise traits could be achieved at the same time by selecting plants with higher lysine content. The relationships between methionine and leucine/isoleucine/ phenylalanine/threonine/lysine contents were significant positive, suggesting that the plant with higher contents of these 5 amino acids might be obtained in rapeseed breeding. In addition, significant positive rP and rG were found between leucine and isoleucine contents, phenylalanine and threonine contents. Negative rP was found between valine and lysine contents. Therefore, the increasing for valine content in rapeseed meal would inevitably lead to a decrease for the content of lysine.The predicted values of genetic effects for 9 parents showed that there existed large difference in genetic main effects at 5 developmental times across environment. Regarding selection of the better parents, Eyouchangjia for lysine, leucine, isoleucine, glutamic acid and glycine contents, Gaoyou 605 for lysine, leucine, isoleucine, phenylalanine, threonine and glycine contents, Zhongyou 821 for phenylalanine and glutamic acid contents, Youcai 601 for leucine, isoleucine, threonine, glutamic acid and glycine contents, Huashuang 3 for leucine, threonine, glutamic acid and glycine contents, could be used because of their stable positive value at mature time across environment. In general, Eyouchangjia was beneficial for improving most essential amino acid content in rapeseed meal of progeny.
|
PDF Full Text Request