Growth curves for Bergamasca sheep in the Brasilia region

Fifteen years of data from a flock of Bergamasca sheep weighed every two weeks was anal zed using non-lineary methods to determine parameters of three growth curves (Brody, Richards and Logistic). The curve parameters were analyzed using Statistical Analysis System (SAS) procedures GLM and CORR to determine factors which influenced them and correlations between the parameters and weights at determined ages. Fixed effects in the model included lambing number, sex of lamb, type of birth and month and age of lambing. The management system is semi-extensive, with the sheep at pasture during the day and imprisoned at night to protect them against predators. Means of the parameters were: Richards; A = 50.024, B = 0.647, k = 0.003, M = 17.771 and R-2= 0.838; Brody; A= 52.763, B=0.885, k=0.003, R-2= 0.842; Logistic; A =47.894, B=-3.895, k=0.007 and R-2= 0.840. The environment influenced the growth of sheep in the region of Brasilia. The Logistic curve estimated adult weights lower than the other curves but the actual weights were better correlated with those estimated by this curve. There was no difference between the curves in terms of R-2, and the error mean square was higher for the Richards curve. The negative correlation between A and k indicates that early maturing animals were less likely to reach high mature weights, compared with late maturing ones. The recommended curve for adjusting growth curves in Bergamasca sheep is the Logistic curve.