肉牛胃肠道甲烷排放模型估算精度的评估分析
发布时间：2018-11-10 11:32
【摘要】：本研究旨在评估6个经典的肉牛胃肠道甲烷排放估算模型的预测精度,分析影响模型预测精度的原因。在湖南望城肉牛养殖场选用17头体况良好的湘中黑牛,分两阶段测定了肉牛体重、营养组分采食量及胃肠道甲烷排放量。本研究选择6个经典的肉牛胃肠道甲烷排放估算模型,包括:以干物质采食量(DMI)为核心参数的估算模型1[CH4(MJ·d~(-1))=1.246×DMI(kg·d~(-1))+0.996]和模型2[CH4(MJ·d~(-1))=-2.07+2.636×DMI(kg·d~(-1))-0.105×DMI2(kg·d~(-1))];以纤维摄入量为核心参数的估算模型3[CH4(MJ·d~(-1))=5.58+0.848×NDF(kg·d~(-1))]和模型4[CH4(MJ·d~(-1))=3.41+0.520×DMI(kg·d~(-1))-0.996×ADF(kg·d~(-1))+1.15×NDF(kg·d~(-1))];以总能摄入量为核心参数的估算模型5[CH4(MJ·d~(-1))=0.065×GEI(MJ·d~(-1))]和模型6[CH4(MJ·d~(-1))=0.081×GEI(MJ·d~(-1))-0.024]。利用预测误差均方(Mean squared prediction error,MSPE)和一致性相关系数(Consistent correlation coefficient,CCC)两种分析方法评估6个估算模型预测肉牛胃肠道甲烷排放量的精度以及影响模型估算精度的原因。结果表明,模型5(CCC=0.86)的估测精度最高,模型1(CCC=0.74)和6(CCC=0.79)次之,模型2(CCC=0.66)、3(CCC=0.22)和4(CCC=0.54)的估算精度最低;模型1和2的估算误差主要来自于整体偏差的偏离(分别为48.8%和70.3%);模型3的估算误差主要来自于回归斜率的偏离(47.6%);模型4的偏差主要来自于整体偏差(29.2%)和回归斜率偏离(28.6%)。IPCC(2006)Tier2推荐的以总能GEI为单一变量的模型5是本试验中6个估算公式预测精度最高的模型。
[Abstract]:The purpose of this study was to evaluate the prediction accuracy of six classical models for estimating methane emissions from gastrointestinal tract of beef cattle, and to analyze the reasons that affect the prediction accuracy of the models. A total of 17 black cattle were selected from Wangcheng Beef Farm in Hunan Province. The body weight, nutrient intake and methane emission from gastrointestinal tract were measured in two stages. In this study, six classical models for estimating methane emissions from gastrointestinal tract of beef cattle were selected. Model 1 [CH4 (MJ d-1) = 1.246 脳 DMI (kg d-1] and model 2 [CH4 (MJ d-1) = -2] were used as core parameters of dry matter intake (DMI). 07 2.636 脳 DMI (kg D1-0.105 脳 DMI2 (kg D1)]; Model 3 [CH4 (MJ d ~ (-1) = 5.58 0.848 脳 NDF (kg d ~ (-1)] and model 4 [CH4 (MJ d ~ (-1) = 3.41 0.520 脳 DMI (kg d ~ (-1)] -1)-0.996 脳 ADF (kg D1) 1.15 脳 NDF (kg D1]; Model 5 [CH4 (MJ d-1) = 0.065 脳 GEI (MJ D1] and model 6 [CH4 (MJ d-1) = 0.081 脳 GEI (MJ d-1-0.024]. The prediction error mean square (Mean squared prediction error,MSPE) and consistency correlation coefficient (Consistent correlation coefficient,CCC) were used to evaluate the accuracy of six models for predicting methane emissions from gastrointestinal tract of beef cattle and the reasons influencing the accuracy of the models. The results showed that the estimation accuracy of model 5 (CCC=0.86) was the highest, model 1 (CCC=0.74) and model 6 (CCC=0.79) took the second place, and models 2 (CCC=0.66), 3 (CCC=0.22) and 4 (CCC=0.54) had the lowest accuracy. The estimation error of model 1 and 2 mainly comes from the deviation of the whole deviation (48.8% and 70.3% respectively), the estimation error of model 3 mainly comes from the deviation of regression slope (47.6%). The deviation of model 4 mainly comes from the global deviation (29.2%) and regression slope deviation (28.6%). IPCC (2006). Model 5, with total energy GEI as a single variable, recommended by Tier2, is the model with the highest prediction accuracy of the six estimation formulas in this experiment.
【作者单位】： 湖南农业大学动物科学技术学院;湖南畜禽安全生产协同创新中心;中国科学院亚热带农业生态研究所;
【基金】：国家自然科学基金项目(31561143009;31472133) 国家科技计划项目(2016YFD0500504) 湖南省科技计划项目(2015WK3043) 中国科学院青年促进会项目
【分类号】：S823.92
[Abstract]:The purpose of this study was to evaluate the prediction accuracy of six classical models for estimating methane emissions from gastrointestinal tract of beef cattle, and to analyze the reasons that affect the prediction accuracy of the models. A total of 17 black cattle were selected from Wangcheng Beef Farm in Hunan Province. The body weight, nutrient intake and methane emission from gastrointestinal tract were measured in two stages. In this study, six classical models for estimating methane emissions from gastrointestinal tract of beef cattle were selected. Model 1 [CH4 (MJ d-1) = 1.246 脳 DMI (kg d-1] and model 2 [CH4 (MJ d-1) = -2] were used as core parameters of dry matter intake (DMI). 07 2.636 脳 DMI (kg D1-0.105 脳 DMI2 (kg D1)]; Model 3 [CH4 (MJ d ~ (-1) = 5.58 0.848 脳 NDF (kg d ~ (-1)] and model 4 [CH4 (MJ d ~ (-1) = 3.41 0.520 脳 DMI (kg d ~ (-1)] -1)-0.996 脳 ADF (kg D1) 1.15 脳 NDF (kg D1]; Model 5 [CH4 (MJ d-1) = 0.065 脳 GEI (MJ D1] and model 6 [CH4 (MJ d-1) = 0.081 脳 GEI (MJ d-1-0.024]. The prediction error mean square (Mean squared prediction error,MSPE) and consistency correlation coefficient (Consistent correlation coefficient,CCC) were used to evaluate the accuracy of six models for predicting methane emissions from gastrointestinal tract of beef cattle and the reasons influencing the accuracy of the models. The results showed that the estimation accuracy of model 5 (CCC=0.86) was the highest, model 1 (CCC=0.74) and model 6 (CCC=0.79) took the second place, and models 2 (CCC=0.66), 3 (CCC=0.22) and 4 (CCC=0.54) had the lowest accuracy. The estimation error of model 1 and 2 mainly comes from the deviation of the whole deviation (48.8% and 70.3% respectively), the estimation error of model 3 mainly comes from the deviation of regression slope (47.6%). The deviation of model 4 mainly comes from the global deviation (29.2%) and regression slope deviation (28.6%). IPCC (2006). Model 5, with total energy GEI as a single variable, recommended by Tier2, is the model with the highest prediction accuracy of the six estimation formulas in this experiment.
【作者单位】： 湖南农业大学动物科学技术学院;湖南畜禽安全生产协同创新中心;中国科学院亚热带农业生态研究所;
【基金】：国家自然科学基金项目(31561143009;31472133) 国家科技计划项目(2016YFD0500504) 湖南省科技计划项目(2015WK3043) 中国科学院青年促进会项目
【分类号】：S823.92
【参考文献】
相关期刊论文 前3条
1 谢天宇;王敏;王荣;颜志成;石惠宇;高民;谭支良;;奶牛胃肠道甲烷排放模型估算精度的评估分析[J];畜牧兽医学报;2015年09期
2 王荣;邓近平;王敏;王玉诗;张玉茹;颜志成;谭支良;;基于IPCC Tier 1层级的中国反刍家畜胃肠道甲烷排放格局变化分析[J];生态学报;2015年21期
3 桑断疾;董红敏;郭同军;罗永明;刘，
本文编号：2322361
本文链接：https://www.wllwen.com/yixuelunwen/dongwuyixue/2322361.html
