近红外漫反射光谱法快速测定秸秆青贮饲料成分含量

Feasibility of Using Near-Infrared Reflectance Spectroscopy for the Analysis of Compositions of Straw Silage

采用傅里叶近红外漫反射光谱技术,结合偏最小二乘回归法,以158个不同种类的秸秆青贮饲料样品建立了常规化学成分和发酵成分含量的近红外定量分析校正模型。其中常规化学成分中,粗蛋白、中性洗涤纤维、酸性洗涤纤维、半纤维素、干物质、粗灰分和酸性洗涤木质素含量的校正模型决定系数R2分别为0.95,0.90,0.86,0.91,0.86,0.95和0.90;发酵成分的pH、乳酸、乙酸、丙酸、丁酸和氨态氮含量的校正模型决定系数R2分别为0.98,0.83,0.85,0.3,0.90和0.92。研究发现除乙酸、丙酸和丁酸之外,其他成分含量的相对分析误差RPD(SD/SECV)均大于2.5;除乳酸、乙酸、丙酸和丁酸之外,所建近红外定量分析模型对独立检验集样品其他成分含量的预测值与化学值的相关决定系数R2均大于0.08。该研究结果对青贮饲料品质的快速检测具有重要的实际意义。

Near-infrared reflectance spectroscopy （NIRS） calibrations of chemical composition in 158 straw silage samples were developed by means of partial least-squares （PLS） regression, Results showed that the correlation coefficients of calibration （R^2） were 0. 95, 0.90, 0. 86, 0. 91, 0. 86, 0.95 and 0. 90 for crude protein, neutral detergent fibre, acid detergent fibre, hemicellulose, dry matter, crude ash and acid detergent lignin respeetively; the R^2 of pH, lactic acid, acetic acid, propionic acid, butyric acid and ammonia were 0. 98, 0.83, 0.85, 0. 36, 0. 90 and 0. 92 respectively. The RPD （SD/SECV） of these parameters were all greater than 2.5 except acetic acid, propionic acid and butyric acid, and the correlation coefficients of validation （R^2） of the parameters were all greater than 0. 80 except lactic acid, acetic acid, propionic acid and butyric acid. These results are of great practieal importance in rapid evaluation of silage quality.