Application Research On Near-infrared Reflectance Spectroscopy For The Analysis Of Ruminant Feed

The aim of this study is to build near-infrared quick analysis model of Inner Mongolia ruminant animals and explore its feasibility by near-infrared reflectance spectroscopy technique.The near-infrared reflectance spectroscopy (model 8620) and principal component regression were applied to build calibration model of conventional chemical composition and TDN content in forage feed sample. The result showed that correlation coefficient of calibration model of DM, CP, EE, Ash, NDF, ADF, ADL and TDN content was 0.7816, 0.9219, 0.8221, 0.9264, 0.9744, 0.9559, 0.9084 and 0.9012, respectively; Validation correlation coefficient was 0.6470, 0.9305, 0.7026, 0.9172, 0.9743, 0.9474,0.8500 and 0.9225. Also, RPD(SD/SEP)was more than 2.5 except of DM and EE content. So, near-infrared prediction model was feasible to quickly measure conventional chemical composition and TDN content in forage feed sample.The calibration model of conventional chemical composition and TDN content in cow concentrate supplement sample was builded by the same methods. The result showed that correlation coefficient of calibration model of CP, EE, Ash, NDF, ADF, ADL and TDN content was0.9132, 0.9016, 0.922, 0.9171, 0.8928, 0.7083 and 0.8346, respectively;Validation correlation coefficient was 0.9042, 0.9035, 0.9075, 0.9138, 0.8801, 0.7939 and 0.9027. Also, RPD(SD/SEP)was more than 2.5 except of ADL. So, near-infrared prediction model was feasible to quickly measure conventional chemical composition and TDN content in cow concentrate supplement sample.Using the same methods, calibration model of conventional chemical composition and index of sensory evaluation in crude forage sample was builded. The result showed that correlation coefficient of calibration model of CP, EE, Ash, NDF, ADF, ADL and TDN, GE, ME, IVDMD, GI was 0.9656, 0.8753, 0.9210, 0.9590, 0.9492, 0.8566, 0.9009, 0.8164, 0.8505, 0.8853, 0.9182, respectively. Validation correlation coefficient was 0.9657, 0.8617, 0.9003, 0.9730, 0.9461, 0.8440, 0.9042, 0.8516, 0.8978, 0.8902 , 0.9082. Also, RPD(SD/SEP)were more than 2.5 except of EE, ADL. So, near-infrared prediction model was feasible to quickly measure index of crude forage sample except of EE, ADL.The calibration model of fermentation parameters in corn silage by the same methods. The result showed that correlation coefficient of calibration model of CP, PH, NH3-N, lactic acid, Acetic acid and Propionic acid content was0.8345, 0.8518, 0.8878, 0.8816, 0.7146 and 0.3425;respectively. Validation correlation coefficient was 0.8861, 0.8956, 0.9141, 0.8247, 0.6528 and 0.4592. Also, RPD(SD/SEP)of CP, PH, NH3-N content was more than 2.5. So, near-infrared prediction model was feasible to quickly measure conventional chemical composition and TDN content in corn silage.