应用近红外(NIRS)技术分析小样品油菜籽粒含油量及硫甙含量

Analysis of glucosinolates content in small rapeseed samples by using the Near Infrared Reflectance Spectroscopy (NIRS) technology

【目的】对193份甘蓝型油菜籽粒进行常量样品(3.0g)及小量样品(0.3g)含油量及硫甙含量近红外光谱分析技术(NIRS)分析和回归模型研究,以期在育种过程中建立实用的油菜小样品含油量及硫甙含量测定数学模型。【方法】利用NIRS对甘蓝型油菜籽粒进行含油量及硫甙含量测定,并进行回归分析。【结果】在众多回归模型中,对于小量样品籽粒含油量及硫甙含量,二次模型均为最优拟合模型,回归方程式分别为:y=40.190+0.892x-0.007x2(R2=0.970);y=-92.040+0.748x+0.002x2(R2=0.960)(y为常量测量结果,x为小量测量结果)。【结论】获得的二次回归方程式可以很好地将小量样品测定数据转化为常量样品分析结果,为高含油量和低硫甙油菜品种的选育奠定基础。

【Objective】In this study, the oil content and glucosinolate content of 193 Brassica napus seed samples（3.0 g）and small samples （0.3 g） were analyzed through the NIRS technology and their regression models were also studied in order to establish mathematical models for the practical detection of oil content and glucosinolate content in small rapeseed samples. 【Method】The oil content and glucosinolate content in Brassica napus seeds were determined by utilizing the NIRS technology, and then the SPSS 18.0 software was used to carry out the regression analysis. 【Result】Within the many regression models for determining the oil content and glucosinolate content in small rapeseed samples, the quadratic model was the best fitting model, and the regression equation was respectively y=40.190＋0.892x-0.007x2（R2=0.970）and y=-92.040＋0.748x＋0.002x2（R2=0.960）（y represents the constant measurement results, x represents small figures measurement results）. 【Conclusion】The quadratic regression equation could effectively transfer the small sample data results into constant conversion data analysis results, therefore able to lay the foundation for breeding Brassica napus varieties with high oil content and low glucosinolate content.