Hermansen J E
National Institute of Animal Science, Department of Research in Cattle and Sheep, Tjele, Denmark.
J Dairy Sci. 1995 Apr;78(4):872-9. doi: 10.3168/jds.S0022-0302(95)76700-5.
The objective of this work was to predict changes in milk fat composition caused by differences in dietary fat. Twenty-two references describing 35 experiments and 108 treatments were used in the analysis. For lauric, myristic, and palmitic acids in milk, proportions in the dietary fat and the total dietary fat concentration were important predictors for their concentrations in milk as well as for stearic and oleic acids in milk. Using a model that included these four parameters, the residual standard deviation around the observed versus predicted line within experiments was approximately 10% of the mean for short-chain fatty acids (< C12); for lauric, myristic, palmitic, and oleic acid; and for total C18 fatty acids in milk. The model also effectively predicted milk fatty acid profile with respect to lauric, myristic, palmitic, and oleic acid and total C18 fatty acids across experiments despite differences in breed, basal diet, and milk yield among experiments. The content of short-chain acids, stearic acid, and poly-unsaturated fatty acids were less effectively predicted across experiments. Possible explanations for the differing predictabilities for different milk fatty acids are discussed.
这项工作的目的是预测由膳食脂肪差异导致的乳脂肪组成的变化。分析中使用了22篇参考文献，描述了35项实验和108种处理方法。对于牛奶中的月桂酸、肉豆蔻酸和棕榈酸，膳食脂肪中的比例以及膳食脂肪总浓度是其在牛奶中的浓度以及牛奶中硬脂酸和油酸浓度的重要预测指标。使用包含这四个参数的模型，实验中观察值与预测值线周围的残差标准差约为短链脂肪酸（<C12）、月桂酸、肉豆蔻酸、棕榈酸和油酸以及牛奶中总C18脂肪酸平均值的10%。尽管不同实验之间在品种、基础日粮和产奶量方面存在差异，但该模型在不同实验中也能有效地预测牛奶中月桂酸、肉豆蔻酸、棕榈酸、油酸和总C18脂肪酸的脂肪酸谱。跨实验对短链酸、硬脂酸和多不饱和脂肪酸含量的预测效果较差。文中讨论了不同乳脂肪酸预测能力不同的可能原因。
