Experimental data (n = 1567) were statistically analysed regarding the impa
ct on milk urea content of dairy cows, using nutritional and animal factors
.
In partial correlation analysis a week stochastic relation between milk ure
a content and selected parameters for the description of rumen metabolism w
as found (r - 0.3 to 0.5). The highest correlation to milk urea content was
found when digestible crude protein intake was related to energy intake (r
= 0.5). Regression analysis showed a significant influence of breed and an
imal, but protein and energy supply of rumen microbes was most important. A
dditionally, stage of lactation, milk protein yield, supply of utilizable p
rotein in duodenum (nXP) and feed intake proved to be significant. The lowe
st residual standard error was found when using digestible carbohydrate and
digestible crude protein intake for describing the rumen nitrogen metaboli
sm in regression analysis. Nevertheless the relative high residua standard
error of 3.9 mg/100 mi of milk urea shows, that a remarkably high proportio
n of variation is caused by factors not considered in the model. This corre
sponds to numerous data in literature. When milk urea data are interpreted,
this inaccuracy should be taken into account.