用近红外反射光谱测定小样本糙米粉的品质性状

比较了用近红外反射光谱（NIRS）技术直接测定约0.5g糙米粉和约3g精米粉样本的表观直链淀粉含量(AAC)、糊化温度（碱消值）和蛋白质含量的效果。结果发现，用精米粉和糙米粉样本测定蛋白质含量均有很好效果，建模时标准误很小而决定系数高，校正标准误（SEC）、检验工作标准误（SEP）、校正决定系数（R#+2）以及检验决定系数（RSQ）依次分别为0.11和0.19,0.20和0.21,0.995和0.932以及0.99和0.91。在建立AAC回归方程时，小样本糙米粉要比精米粉差一些，前者的SEC、SEP、R#+2和RSQ分别为0.933、2.07、0.927、和0.65，而用精米粉时依次为0.426、0.91、0.985和0.93。用小样本糙米粉测定糊化温度（碱消值）时，SEC、SEP、R#+2和RSQ分别为0.349、0.48、0.827和0.74，而用精米粉时依次为0.224、0.36、0.949和0.84。在对水稻育种过程中用小样本糙米粉进行稻米品质的NIRS测定作了讨论。

Analysis of Grain Quality Characters in Small Ground Brown Rice Samples by Near Infrared Reflectance Spectroscopy

The effectiveness of near infra red reflectance spectroscopy (NIRS) analysis was compared between ground milled rice (GMR) samples ( ca . 3g) and ground brown rice (GBR) samples ( ca . 0.5g) for protein content, apparent amylose content (AAC) and gelatinization temperature ( alkali spreading value) of rice grains. The results were reported as follows. For protein content, good regression equations could be established both for GMR and GBR. The standard error for ca libration (SEC) and validation performance (SEP), the determination coefficients for ca libration (R 2) and validation (RSQ) for GMR and GBR samples were 0.11 and 0.19, 0.20 and 0.21, 0.995 and 0.932, and 0.99 and 0.91, respectively. For AAC ca libration, GBR samples were less effective than GMR ones, the SEC,SEP,R 2 and RSQ for GBR samples was 0.933, 2.07, 0.927 and 0.65, respectively, and those for GMR samples was 0.426, 0.91, 0.985 and 0.93, respectively. For alkali spreading value, the SEC,SEP,R 2 and RSQ was 0.349, 0.48, 0.827 and 0.74 for GBR, respectively, and 0.224, 0.36, 0.949 and 0.84 for GMR, respectively. The appli ca tion of NIRS analysis by using small ground brown samples for rice grain quality evaluation in breeding programs was discussed.