Rapid Analysis of Main Quality Parameters in Forage Soybean by Near-Infrared Spectroscopy
JIANG Yan1, 2, MENG He1, ZHAO Yi-rong1, WANG Xian-xu1, WANG Sui1, XUE En-yu3, WANG Shao-dong1*
1. Northeast Agricultural University/National Soybean Engineering Technology Research Center, Harbin 150030, China
2. Heilongjiang Academy of Green Food Science, Harbin 150028, China
3. Heilongjiang Rural Industry Development Center, Harbin 150036, China
Abstract：Forage is the material basis of animal husbandry production. The detection and evaluation of the nutritional value of forage raw materials and feed products are an important link in feed production. Facing the situation of low crude protein content in forage resources and relying on many imported feeds, soybean, as a high-quality, high protein legume forage, is an important resource for animal husbandry production and utilization. The feeding quality parameters of different forage soybean and different cutting periods can evaluate the feeding performance of forage soybean. However, the chemical method is mainly used for detection, which is cumbersome, long test cycle and is easy to cause operation error. Moreover, the rapid detection of main feeding quality indexes of forage soybean is still blank, which needs to be developed and utilized urgently. Because of the wide application of near-infrared spectroscopy in detection and feed analysis, the whole plant samples of different soybean varieties in different cutting periods were collected by near-infrared spectroscopy in the range of 950～1 650 nm. The content of crude protein (CP), neutral detergent fiber (NDF) and acid detergent fiber (ADF) was detected according to the national standard or industry standard chemical method. The 150 samples data were divided into calibration and verification set according to 3∶2. The prediction models of three main quality parameters CP, NDF and ADF content of forage soybean, were established by combining one or more of four different spectral pretreatment methods, including first-order derivative (NW1st), second-order derivative (NW2nd), standard normal variable transformation (SNV) and detrending (DE-trending), and partial least squares (PLS) regression algorithm. By comparing the coefficient of determination (R2) and root mean square error (RMSE) of calibration set and validation set in regression models, the results showed that the model established by NW1st+DE-trending+SNV+PLS had the best effect. The R2C and R2P of the calibration set and validation set in forage soybean CP content model were 0.96 and 0.95 respectively, the R2C and R2P of NDF content model were 0.90 and 0.89 respectively, and the R2C and R2P of ADF content model were 0.94 and 0.93 respectively. The accuracy and stability of the model were further confirmed by the test and analysis of the validation, and rapid analysis of near-infrared spectroscopy (NIRS) method for the qualitative detection of forage soybean was formed. With the increase of forage soybean quality parameter data, the quality detection model of forage soybean will continuously improve. This method expands the detection category and range of forage resource quality by near-infrared spectrometer and is accurate and efficient, which is conducive to the development and effective utilization of high-quality high protein forage resources.
[1] Ferreira D S, Pallone J A L, Poppi R J. Food Research International, 2013, 51: 53.
[2] Leite D C, Corrêa A A P, Júnior L C C, et al. Journal of Food Composition and Analysis, 2020, 91: 103536.
[3] Guo W L, Du Y P, Zhou Y C, et al. World Journal of Microbiology and Biotechnology, 2012, 28(3): 993.
[4] LIU Yan-de, XU Hai, SUN Xu-dong, et al(刘燕德, 徐 海, 孙旭东, 等). Spectroscopy and Spectral Analysis(光谱学与光谱分析), 2020, 40(3): 922.
[5] Harris P A, Nelson S, Carslake H B, et al. Journal of Equine Veterinary Science, 2018, 71: 13.
[6] Modroño S, Soldado A, Martínez-Fernández A, et al. Talanta, 2017, 162: 597.
[7] Xu R, Hu W, Zhou Y, et al. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, 2020, 224: 117400.
[8] Brogna N, Palmonari A, Canestrari G, et al. Dairy Science, 2017, 101: 1234.
[9] Righi F, Simoni M, Visentin G, et al. Livestock Science, 2017, 206: 105.
[10] REN Zhi-shang, PENG Hui-hui, HE Zhuang-zhuang, et al(任志尚, 彭慧慧, 贺壮壮, 等). Transactions of the Chinese Society for Agricultural Machinery(农业机械学报), 2020, 51(S2): 466.
