以常规成分为变量估测饲料中必需氨基酸含量的初步探讨

本文探讨了玉米、小麦、高粱、豆饼(粕)、菜籽饼(粕)、花生饼(粕)、国产鱼粉、小麦麸、米糠、米糠饼、米糠粕、苜蓿草粉、大豆、黑大豆、蚕豆、豌豆的粗蛋白、粗纤维、粗脂肪和粗灰分与必需氨基酸的多元回归方程和粗蛋白与必需氨基酸之间的一元回归方程。一元回归方程估测误差(RSD)略高于多元回归方裎估测误差,但相关程度两者基本一致。方差分析表明,玉米、小麦、花生饼(粕)、国产鱼粉、小麦麸、米糠、苜蓿草粉、黑大豆、大豆、蚕豆、豌豆的常规成分与必需氨基酸含量存在显著的相关关系(P<0.05);大麦、高粱、米糠饼(粕)和菜籽饼(粕)常规成分与苯丙氨酸、组氧酸、精氨酸等必需氨基酸存在显著扣关关系(P<0.05),与赖氨酸、亮氨酸、蛋氨酸等必需氨基酸不存在相关关系(P>0.05);豆饼(粕)常规成分与必需氨基酸不存在相关关系。利用朱参加计算的玉米和小麦麸样品数据资料比较了粗蛋白与必需氨基酸之间的一元回归方程估测值、按粗蛋白递增比例计算值和实测值之间的差异。表明用回归方程估测较按粗蛋白质递增比例推算,更接近于实际测定值,具有显著相关关系的回归方程应用于生产以估测某些必需氨基酸含量是可行的。

THE PREDICTION OF ESSENTIAL AMINO ACID CONTENTS FROMTHE APPROXIMATE ANALYSIS OF FEEDSTUFFS IN CHINA

The regression equations of corn, wheat, sorghum, soybean meal, rapeseed meal, peanut meal, fish meal (China), wheat bran, rice bran, rice bran (expeller or solvent), alfalfa meal, soybean (full fat), black soybean, horsebean, peas are established between crode protein, crude fibre, crude fat, crude ash and the essential amino acids such as are Thr, Val, Met, Cys, ne, Leu, Phe, Lys, His, Arg. The regression equations of the feeds concerned are also established between crude protein and the essential amino acids. The resiclual standard deviations of the bivariate equations is greater than those of thd mutiple equations. The results of variance analysis show that the multipleand bivariate equations of corn t wheat, peanut meal, fish meal (Product of china), wheat bran, black soybean, soybean, horsebean, peas, alfalfa meal reach significant level for all the essential amino acids concerned (P<0.05). Those of barley, sorghum, rice bran (expeller or solvent), rapeseed meal reach significant level for some of the essential amino acids (P<0.05), but for the others don't (P>0.05). By comparing the lysine contents of samples from corn and wheat bran which are not used in regression equation calculations with the values based on the corresponding blvariate equations and the lysine proportions to crude protein, it is found that thd regression equations can predict the lysine contents of the samples concerned. It is suggested that the bivariate equations could be applied to pratical situations.