几种利用乳中尿素氮浓度估测尿氮排泄量模型的评价

本文利用34组国内外报道的荷斯坦牛乳中尿素氮浓度和尿氮排泄量的实测数据，对目前提出的部分利用MUN估测尿氮排泄量的模型进行了比较。结果表明，Zhai（2005）提出的模型UN（g/d）：10．1×MUN（mg／dL）＋47.3]预测效果较好（P〉0．05）。划分MUN浓度范围分别建立模型可能会提高预测的准确度。此外，根据34组数据做简单回归：UN（g/d）=12．78×MUN（mg/d1）＋28．15（n=34，R2=0．59）。

Evaluation of Some Models to Predict Urea Nitrogen Excretion Using Milk Urea Nitrogen Concentration

34 groups of statistics which included milk urea nitrogen（MUN） concentration and excetion of urea nitrogen （UN） were used to evaluate accuracy of some models, by which excretion of UN was estimated by MUN concentration. The results concluded that excretion of UN calculated by MUN concentration in Zhai＇s（2005） model UN（g/d）=10.1 ×MUN（mg/dl）＋47.3] was more similar with true result（P〉0.05）. Accuracy may increase when different modles summarized according to different ranges of MUN concentration. Besides, simple regression was made by those stastistics： UN（g/d）=12.78×MUN （mg/dl）＋28.15 （n=34, R2=0.59）.