Silva, E. P. [UNESP]Malheiros, E. B. [UNESP]Sakomura, N. K. [UNESP]Venturini, K. S. [UNESP]Hauschild, L. [UNESP]Dorigam, J. C. P. [UNESP]Fernandes, J. B. K. [UNESP]2015-10-212015-10-212015-03-01Livestock Science, v. 173, p. 69-77, 2015.1871-1413http://hdl.handle.net/11449/129221This experiment aimed to model the responses of laying hens to lysine intake (Lys) using three mathematical models. The design used was completely randomized, with eight treatments, six repetitions with eight hens per experimental unit, and three periods of 28 days for each period, using 384 Dekalb White hens. Statistical analyses were performed for treatments, repetitions and period using the procedure of repeated measures in time. The responses for EM and FCR for the different intakes of Lys were adjusted using the models broken line (BL), quadratic polynomial (QP) and reading model (RM). There was an effect of Lys levels and period and an interaction between the levels of Lys and period. The models were adjusted for each experimental period. The biological optimal level of Lys intake was obtained using a combination of the models BL and QP, which was close to the estimate of RM of 707, 660 and 669 mg/bird per day, which allows to meet the requirement of approximately 97% population in the periods of 37-40, 41-44 and 45-48 weeks old. For the cost of L-Lysine HCl and for egg price variation, the economic optimum intake was estimated at 699, 658 and 667 mg of Lys/bird per day for the referred periods. (C) 2015 Elsevier B.V. All rights reserved.69-77engAmino acidsBroken linePopulationQuadratic polynomialReading modelArtigo10.1016/j.livsci.2015.01.005WOS:000350517900010Acesso restrito472659619394902261523290002748581612969183171944