节点文献
水稻蛋白质近红外模型的建立及其在育种中的应用研究
Establishment of Math Models of NIRS Analysis for Protein Content in Seed of Rice and Application in Rice Breeding (Oryza Sativa L.)
【作者】 李君霞;
【作者基本信息】 中国农业大学 , 作物遗传育种, 2005, 硕士
【摘要】 稻米品质改良是目前水稻研究中的热点和难点问题。氮素代谢在植物的新陈代谢中占主导地位,快速测定其含量,对研究植物氮素吸收、运输和代谢规律,确定农产品品质、营养价值等具有一定的意义。文章借助近红外技术,建立快速高效的稻米蛋白质含量测定技术分析平台,并应用于育种实践,为下一步优异基因的挖掘和特异功能稻的研发提供了操作基础。本文主要(1)以191份水稻糙米样品为材料,研究近红外漫反射光谱测定水稻糙米蛋白质方法和可行性(2)筛选北方粳稻种质资源,研究不同环境下、不同遗传背景糙米蛋白质含量效应(3)蛋白质含量高的种质资源和农艺性状好的品种配制组合,进行杂交组合后代蛋白质含量的遗传、分离、变异、与单株产量间的相关性分析,结果如下: 1.借助偏最小二乘法(PLS)分析191份糙米样品,采用优化模型的目标函数R/(1+RMSECV),对近红外PLS定量分析模型进行优化研究,目标函数作为模型优化目标的判定,综合了模型的两个重要参数,使评判标准更加客观准确,克服了以往同类研究中单一参数评价的弊端,结果表明:最优光谱数据预处理(一阶导数与多元散射校正)函数值0.73、最佳谱区的目标函数值0.74(谱区为11998.9cm-1-5449.8cm-1和4601.3cm-1-4246.5cm-1),说明目标函数不失为模型优化的一个重要的参考指标;比较各种不同的光谱预处理方法和谱区选择的效果,说明采用一阶导数和矢量标准化预处理、谱区为11998.9cm-1-5449.8cm-1和4601.3cm-1-4246.5cm-1建立的校正模型检验和预测效果最佳,其决定系数为0.9886,相对标准偏差RSD为2.1%,表明近红外反射光谱这种简单、方便、快速的分析技术,非常适用于低世代育种材料的筛选和水稻糙米蛋白质含量的快速测定。 2.筛选688份优异种质资源,研究不同环境、不同磷水平糙米蛋白质含量变化,结果表明:测定的粳稻种质资源蛋白质含量变幅6.34-19.63%,平均含量为11.75%,含量在16%以上的稻种共14份,占总数2%,可供水稻品质育种工作者在以后的高蛋白质育种中选择利用,筛选出低于7%品种3份,占总数0.4%,可供培育低含量有特殊作用功能性的水稻品种选用;不同来源和类型的品种蛋白质含量存在广泛差异,不同品种间差异极显著、环境间差异显著、磷处理差异不显著,蛋白质含量的变化与不同磷水平处理下生物产量的增减有相同的趋势,说明水稻植株氮代谢与磷元素的吸收存在极显著正相关。 3.分析水稻蛋白质育种中杂交组合后代蛋白质含量的遗传、分离、变异、与单株产量间的相关性,结果表明:水稻蛋白质含量F2到F3代广义遗传率不高于30%;杂交组合后代籽粒蛋白质含量,单株分布呈单峰连续变异,这也证明了水稻蛋白质含量是多基因控制的数量性状,这与前人的研究结果是一致的;蛋白质含量与单株产量存在显著的负相关,说明期望获得兼顾稻米产量与蛋白质含量两个目标性状的优良品种有一定的难度。研究中分析筛选优异种质的杂交后代,蛋白质含量的母体效应比较明显,提出了粳稻蛋白质育种的改良策略:以蛋白质含量高的材料做母本,产量高、农艺性状好的品种作父本,可以增加优质基因筛选的机率,加快育种进程。
【Abstract】 The objectives of this study were: (1) investigate the feasibility and methods of measuring protein content of rice kernel with near infrared reflectance spectroscopy (NIRS) by using 191 brown rice (5.9%-14.5%) (2) by using the model, screen the optimum rice resource germplasm with protein content, study the effect of protein content under different phosphorus treatments and backgrounds (3) study the heredity, variance, separation and the correlation between protein content and yield of per plant with some crosses between promising rice varieties with high yield and good quality and rice resource germplasm with high protein content. The results were as follows:1. PLS was used to determine protein content of 191 rice samples (5.9%-14.5%) by different wavelength ranges and data preprocessing procedures chosen for regression and information extraction, and it’ s object function [R/(1+RMSECV)] is structured. The method is applied to the research of the calibration model, and a satisfying result is reached. The results demonstrated that the method of selecting advantageous wavelength ranges and data preprocessing procedures is feasible to obtain significant prediction precision, regarding maximizing R/(1+RMSECV) as the object function. This model is more representative, stable and robust . The optimal model was developed. This model’s determination coefficient (Rcal2) and relative standard deviation (RSD) were 0.9886 and 0.021,respectively. The model was showed it was significant correlation and lower error between near-infrared value and true value, and is lower than 0.4.So,it was showed that NIRS is feasible to determine protein content of rice kernels.2. 688 rice cultivans in China Oryza sativa germplasm and some lines from IRRI showed average protein content of 11.75%, ranged 6.34-19.63%, among the materials tested 14 lines were found higher than 16%, 3 lines lower than 7%,they will be used in high protein and functional nutrition rice breeding. Varieties of rice were treated with different concent phosphorus.The results showed that varieties of rice are very different sensitiveness in phosphorus stress,but it is insignificant between different treatments.So,this indicated the significant correlation between nitrogen metabolism and phosphorus absorbance .3. The genetic variability of grain protein content and the correlation between protein content and per plant yield were analyzed. The results showed those populations had wide variation, protein content in rice grain belongs to quantitative traits controlled by polygene.The heritability was low.Significant negative correlation was found between protein content and yield.Parents protein content affects progenies,whereas the female parents have much bigger effect.It is beneficial to use parents with high protein content as female parents in breeding combinations.
【Key words】 Oryza sativa L.; NIRS; protein content; calibration equations; germplasm; inheritance;
- 【网络出版投稿人】 中国农业大学 【网络出版年期】2005年 05期
- 【分类号】S511
- 【被引频次】5
- 【下载频次】406