The main objective of this study was to develop a dynamic energy balance model for dairy goats to describe and quantify energy partitioning between energy used for work (milk) and that lost to the environment. Increasing worldwide concerns regarding livestock contribution to global warming underscore the importance of improving energy efficiency utilization in dairy goats by reducing energy losses in feces, urine and methane (CH4). A dynamic model of CH4 emissions from experimental energy balance data in goats is proposed and parameterized (n = 48 individual animal observations). The model includes DM intake, NDF and lipid content of the diet as explanatory variables for CH4 emissions. An additional data set (n = 122 individual animals) from eight energy balance experiments was used to evaluate the model. The model adequately (root MS prediction error, RMSPE) represented energy in milk (E-milk; RMSPE = 5.6%), heat production (HP; RMSPE = 4.3%) and CH4 emissions (E-CH4; RMSPE = 11.9%). Residual analysis indicated that most of the prediction errors were due to unexplained variations with small mean and slope bias. Some mean bias was detected for HP (1.12%) and E-CH4 (1.27%) but was around zero for E-milk (0.14%). The slope bias was zero for HP (0.01%) and close to zero for E-milk (0.10%) and E-CH4 (0.22%). Random bias was >98% for E-CH4, HP and E-milk, indicating non-systematic errors and that mechanisms in the model are properly represented. As predicted energy increased, the model tended to underpredict E-CH4 and E-milk. The model is a first step toward a mechanistic description of nutrient use by goats and is useful as a research tool for investigating energy partitioning during lactation. The model described in this study could be used as a tool for making enteric CH4 emission inventories for goats.

本研究的主要目的是为奶山羊开发一个动态能量平衡模型，以描述和量化用于工作的能量（牛奶）和流失到环境中的能量之间的能量分配。全世界对牲畜对全球变暖的贡献日益关注，这凸显了通过减少粪便、尿液和甲烷 ( CH 4 ) 中的能量损失来提高奶山羊能源效率利用的重要性。提出并参数化了山羊实验能量平衡数据中CH 4排放的动态模型（n = 48 个体动物观察）。该模型包括 DM 摄入量、NDF 和饮食中的脂质含量作为 CH 4排放的解释变量。一个额外的数据集（n = 122 只个体动物）来自八次能量平衡实验，用于评估模型。该模型（根 MS 预测误差，RMSPE）充分代表了牛奶中的能量（E-milk； RMSPE = 5.6%）、产热（HP； RMSPE = 4.3%）和 CH 4排放（E-CH 4；RMSPE = 11.9% ） ）。残差分析表明，大多数预测误差是由于具有较小均值和斜率偏差的无法解释的变化造成的。HP (1.12%) 和 E-CH 4 (1.27%)检测到一些平均偏差，但 E-milk (0.14%) 约为零。HP (0.01%) 的斜率偏差为零，而 E-milk (0.10%) 和 E-CH 4 (0.22%) 的斜率偏差接近零。E-CH 的随机偏差 > 98%图4，HP 和 E-milk，表明非系统性错误，并且模型中的机制被正确表示。随着预测能量的增加，模型倾向于低估 E-CH 4和 E-milk。该模型是迈向山羊营养利用机制描述的第一步，可用作研究泌乳期间能量分配的研究工具。本研究中描述的模型可用作编制山羊肠道 CH 4排放清单的工具。