近红外光谱技术测定紫花苜蓿青贮鲜样的发酵品质

及时了解青贮饲草的发酵品质，如青贮饲草中氨态氮(NH₃-N)，乳酸(LA), 乙酸(AA)，丁酸(BA)含量，对畜牧业生产具有重要意义。应用偏最小二乘回归法(PLS)、傅里叶变换近红外光谱技术和液氮冷冻制样技术，建立了适合不同品种，不同生育期，不同茬次和不同青贮方法即时测定青贮苜蓿鲜样中氨态氮，乳酸，乙酸，丁酸的近红外校正模型，以期对NIRS在测定紫花苜蓿青贮样品中这些物质的可行性进行分析。所建模型的交叉检验决定系数(R2_CV)为0.602 4～0.949 7，交叉检验标准误(RMSECV)为0.559～3.78 g·kg-1鲜重。用检验集样品对模型进行外部检验，预测相关系数(r)为0.882 6～0.985 3，预测标准误为0.571～3.15 g·kg-1鲜重。结果表明，采用适当的样品处理方法和光谱分析技术可以实现近红外光谱技术对苜蓿青贮样品发酵品质的评定。

Evaluation of Fermentation Character of Alfalfa Silage Through Near Infrared Reflectance Spectroscopy (NIRS)

It is very important to evaluate the fermentation character of alfalfa silage using near infrared reflectance spectroscopy technology (NIRS) for animal production, including the content of NH₃-N, lactic acid, acetic acid and butyric acid in silage. In order to evaluate the feasibility of using NIRS to analyze the formation character of alfalfa silage, the near infrared reflectance spectroscopy models were built for NH₃-N, lactic acid, zcetic acid and butyric acid in this experiment. Partial least square regression (PLS), Fourier transform technology and sample preparation with liquid nitrogen technology were used to optimize the model. The analyzed samples were obtained with different cultivars, maturity, cuttings and ensiling method. The determination of cross validation was between 0.602 4 and 0.949 7. The standard errors of cross validation were between 5.59×10-1 and 3.78 g·kg-1 fresh weight. The validation samples were used to test the performance of the models. The correlation coefficients between the chemical value and the NIRS value were between 0.882 6 and 0.985 3, and the root mean square errors of prediction were between 5.71×10-1 and 3.15 g·kg-1 fresh weight. The results showed the NIRS could evaluate the fermentation of the fresh forage.