Use of plasma urea nitrogen as a rapid response criterion to determine the lysine requirement of pigs.

Five experiments were conducted to evaluate the potential use of plasma urea N (PUN) concentrations as a rapid response criterion to determine amino acid requirements. A preliminary experiment (Exp. 1) indicated that a 3-d feeding time was required to re-equilibrate PUN concentrations after a change in the dietary concentration of lysine. In Exp. 2, 3, and 4, PUN was used to estimate the lysine requirement of growing pigs at different specific BW. Thirty individually penned crossbred pigs weighing 32 and 44 kg in Exp. 2 and 3, respectively, were assigned to five dietary treatments (.60, .70, .80, .90, and 1.00% lysine) for 5 d. The PUN decreased quadratically (P < .05) to increasing dietary lysine. A two-slope, broken-line regression model estimated the requirement at .85% in Exp. 2 and at .76% in Exp. 3. In Exp. 4, 60 crossbred pigs (30 barrows and 30 gilts) weighing 70 kg were assigned to five dietary lysine concentrations: .50, .60, .70, .80, and .90% for 4 d. Increasing lysine caused PUN to decrease quadratically (P < .01). The estimated requirements were different (P < .05) between sexes: .69% for barrows and .75% for gilts. In Exp. 5, the validity of using PUN as a rapid response criterion was verified by comparing the estimated lysine requirement based on PUN with the requirement determined in a 7-d N balance. Twenty crossbred barrows averaging 19 kg were used. Dietary lysine concentrations were .60, .75, .90, 1.05, and 1.20%. A quadratic response was observed in PUN (P < .05) and N retention (NR) (P < .01) with increasing lysine. The estimated lysine concentrations that maximized rates of NR and minimized PUN (1.03 vs. 1.05) were not different (P > .10). Therefore, PUN concentrations can be used in short-term trials to accurately estimate the dietary lysine required to maximize total N utilization in pigs at a specific BW. In addition, the two-slope broken-line regression model had the highest R2 and the lowest mean square error compared with three other models as means for estimating lysine requirement from PUN concentrations.