Quantitative Analysis of Herbage Nutritional Components on Natural Pasture by Near Infrared Diffuse Reflectance Spectroscopy
In order to determine the nutrient content of native herbage rapidly in Inner Mongolia Xilin Gol League rangeland,407 samples of main species and the mixed herbage were collected from May to November in 2016.This study utilized the modified partial least squares (MPLS) to get the calibration of crude protein (CP),neutral detergent fiber (NDF),acid detergent fiber (ADF),lignin (ADL),crude ash (Ash),crude fat (EE),calcium (Ca),phosphorus (P) by combing near-infrared reflectance full spectrum scanning technology with the laboratory test data.The results show that the relative error of external verification (RPD (SD/SEP) of Ca,NDF,DM,CP and Ash is greater than 3,the correlation coefficient RSQ between the value predicted by NIRS and chemical method is greater than 0.9,which indicated that calibration of Ca,NDF,DM,CP,Ash is good,the establishment of the calibration model can be used for practical detection;The quantitative analysis of the components by employing NIRS is feasible as the relative error of external verification of ADF is 2.5＜RPD＜3 and RSQ is 0.881.However,the accuracy needs to be further improved.Meanwhile,the prediction model of the elements P,ADL and EE cannot be used for quantitative analysis of near-infrared spectroscopy since the relative error of external verification of the elements is RPD＜2.5,RSQ is 0.56、0.472 and 0.806,respectively.In this paper,the analytical model for the natural herbage contents on the rangeland provides convenient for the rapid test and database of the herbage nutrient.
Near infrared spectroscopyNatural herbageNutritional compositionQuantitative analysis
NETL
NSTL
万方数据
维普
