Statistical analysis
All data were analyzed using SAS (SAS inst. Inc., Cary, NC, USA). In Exp. 1 and Exp. 2, data were analyzed usingthe Proc-Mixed procedure of SAS with each replicate as the experimental unit. The statistical model for the DE, ME and SID of AA values had treatment as fixed effects, period and pig as a random effect. In Exp. 3, all data were analyzed as a randomized complete block design with pen as the experimental unit. Statistical differences among treatments were separated by Student Neuman Keul’s multiple range test. When the ANOVA was significant, differences betweentreatments were tested using Least Square Means (t-tests). Significance was taken at p