牛、羊精料补充料养分含量近红外光谱定标模型构建  

Construction of Near Infrared Spectroscopy Calibration Model for Nutrient Contents of Cattle and Sheep Concentrate Supplements

作　　者：史艳丽 王涛[1] 孙旭春 李飞[1] 刘保仓 邓晓裕 张兆才 翁秀秀[1] SHI Yanli;WANG Tao;SUN Xuchun;LI Fei;LIU Baocang;DENG Xiaoyu;ZHANG Zhaocai;WENG Xiuxiu(Key Laboratory of Grassland Livestock Industry Innovation,Ministry of Agriculture and Rural Affairs,State Key Laboratory of Herbage Improvement and Grassland Agro-Ecosystems,College of Pastoral Agriculture Science and Technology,Lanzhou University,Lanzhou 730020,China;Linxia Hui Autonomous Prefecture Animal Husbandry Technology Extension Station,Linxia 731199,China;Aksu Taikun Feed Co.,Ltd.,Aksu 842008,China;Gansu Runmu Biological Engineering Co.,Ltd.,Jinchang 737100,China)

机构地区：[1]兰州大学草地农业科技学院,草种创新与草地农业生态系统全国重点实验室,农业农村部草牧业创新重点实验室,兰州730020 [2]临夏回族自治州畜牧技术推广站,临夏731199 [3]阿克苏泰昆饲料有限责任公司,阿克苏842008 [4]甘肃润牧生物工程有限责任公司,金昌737100

出　　处：《动物营养学报》2024年第12期8088-8099,共12页CHINESE JOURNAL OF ANIMAL NUTRITION

基　　金：临夏州牛羊产业重大科技支撑项目(KJJC⁃LX⁃2023⁃06);种业攻关和农业科技支撑项目(KJZC⁃2023⁃24)。

摘　　要：本研究旨在利用近红外光谱(NIRS)技术结合改良偏最小二乘法(MPLS)建立牛、羊精料补充料中7种常规营养成分[水分、有机物(OM)、粗蛋白质(CP)、中性洗涤纤维(NDF)、酸性洗涤纤维(ADF)、粗脂肪(EE)和粗灰分(Ash)]和5种矿物元素(钾、磷、钠、镁和铁)含量的快速预测模型。结果表明:1)常规营养成分中,Ash含量预测模型的预测决定系数(R_(P)^(2))和相对预测偏差(RPD)分别为0.74和1.96,仅能用于快速筛选分析;水分、OM、CP、NDF、ADF和EE含量预测模型的R_(P)^(2)和RPD分别为0.90~0.98和2.95~6.66,均可用于实际定量分析。2)矿物元素中,钾和铁含量预测模型的R_(P)^(2)分别为0.70和0.73,RPD分别为1.84和1.83,可用于快速筛选分析;磷、钠和镁含量预测模型的R_(P)^(2)分别为0.56、0.31和0.63,RPD分别为1.50、1.12和1.65,需进一步调整优化才能用于实际生产。综上所述,本研究成功建立了牛、羊精料补充料营养成分及矿物元素含量的NIRS快速预测模型,多数成分预测准确性高,部分矿物元素模型需进一步优化。The aim of this study was to establish the rapid prediction models for the contents of seven conven⁃tional nutrients[moisture,organic matter(OM),crude protein(CP),neutral detergent fiber(NDF),acid detergent fiber(ADF),ether extract(EE)and ash(Ash)]and five mineral elements(potassium,phosphor⁃us,sodium,magnesium and iron)in cattle and sheep concentrate supplements by near infrared spectroscopy(NIRS)combined with modified partial least squares(MPLS).The results showed as follows:1)in conven⁃tional nutrients,the coefficient of determination in prediction(R_(P)^(2))and relative prediction deviation(RPD)of Ash content prediction model were 0.74 and 1.96,respectively,which could be used for rapid screening analy⁃sis;the R_(P)^(2)and RPD of moisture,OM,CP,NDF,ADF and EE content prediction models were 0.90 to 0.98 and 2.95 to 6.66,respectively,which could be used for actual quantitative analysis.2)In mineral elements,the R_(P)^(2)of potassium and iron content prediction models were 0.70 and 0.73,respectively,and the RPD were 1.84 and 1.83,respectively,which could be used for rapid screening analysis;the R_(P)^(2)of phosphorus,sodium and magnesium content prediction models were 0.56,0.31 and 0.63,respectively,and the RPD were 1.50,1.12 and 1.65,respectively,which needed to be further adjusted and optimized for practical production.In conclusion,this study successfully establish NIRS rapid prediction models for the nutrient and mineral element content of cattle and sheep concentrate supplements,most of the component predictions have high accuracy,but some mineral element models need further optimization.

关 键 词：近红外光谱技术 精料补充料 常规养分 矿物元素 快速预测模型 

分 类 号：S816.17[农业科学—饲料科学]