This work had for objective to apply the response plateau technique to evaluate the methodology and the models in the requirements of Zn in broiler. The experiment was a completely randomized design with (2 x 2 x 9) factorial structure, being 2 experiments, 2 Sex (sexes) and 9 levels of Zn, with 8 replications. The levels of Zn were: 0, 15, 30, 45, 60, 75, 90, 105, 120 ppm. The studied variable was Zn level in the tibia in ppm. The models were fitted using the averages of the doses from the initial analysis of variance in SAS PROC NLIN (SAS®). Two nonlinear models (MNLE1 and MLNE2) and the quadratic polinomial model (MPQ), both with plateau, were compared. The three models results equivalent fitting to data. As MNLE1 has the easiest interpretation it was sugested as the best model.
broilers; nonlinear models; response plateau
