statistical analysis of some factors affecting crude protein balance in lactating dairy cows

some factors affecting crude protein balance (cpb), as a measure of efficiency, were studied using 2824 records of 501 lactating holstein cows. the cpb (kg d-1) was calculated as crude protein (cp) intake (kg) minus cp yield (kg). two different statistical methods including principal components (pcs) and general linear model analyses were used to study the effects of different factors. crude protein balance had a moderate correlation with pcs 1 (-0.206), 2 (0.318) and 3 (-0.281), which accounted for 65.6% of total variations. age, parity, lactation stage, pregnancy stage, dry matter intake (dmi), milk yield (my), fat corrected milk yield (fcm), milk fat percentage (f%), milk lactose percentage (l%), milk fat yield (fy), milk protein yield (py), milk lactose yield (ly), dietary levels of net energy for lactation (nel), cp, ruminally undegradable protein (rup), neutral detergent fiber (ndf) and ether extract (ee) and income over feed cost (iofc) were correlated to cpb, at least, in one of pcs 1, 2 or 3. in general, linear model analysis cpb was significantly affected by parity and lactation stage and had significant partial linear regression coefficients on dmi (kg d-1), f%, fy (kg d-1), py (kg d-1), nel (mcal kg-1) and dietary levels of cp (%), rup (%), ndf (%) and ee (%). the quadratic partial regression coefficients of cpb on nel, cp, rup and ndf were also significant. the estimates for optimum dietary levels of nel, cp and rup for minimizing cpb in the studied population were 1.49 mcal/kg, 11.29% and 7.58%, respectively. in comparison to nrc’s estimates, it seems that, more nel, rup and ndf and less cp are needed to minimize protein balance in lactating dairy cows.