*2.5. Statistical Analysis*

All data were analyzed using *Proc* Mixed model of SAS as follows:

$$\mathbf{Y}\_{\rm ijkl} = \mu + \rho\_{\mathbf{i}} + \mathbf{y}\_{\mathbf{j}} + \mathbf{t}\_{\mathbf{l}} + \tau\_{\mathbf{k}} + \varepsilon\_{\mathbf{ijkl}} \tag{1}$$

where Yijkl are the observation parameters, µ is the overall mean, ρ<sup>i</sup> is the random effect of animal, γ<sup>j</sup> is the random effect of periods, t<sup>l</sup> is the random effect of time, τ<sup>k</sup> is the fixed effect of treatments, and εijkl is the error term. Differences among treatments were compared using Duncan's multiple range test, statistically accepted at *p* < 0.05.
