A comparison of non-linear models for describing weight-age relationships in n’dama cattle

Abstract

Three non-linear models were used to fit weight-age data of N’Dama cattle reared under ranching conditions at Fashola, Nigeria. Comparisons were made among these models for computational difficulty, goodness of fit and lack of bias in estimate of mature weight. Brody’s and Bertalanffy’s growth models were easy and took less time to reach convergence in comparison to Richards’ model. Richards’ model with least residual mean squares was the best fit to the observed growth pattern of male and female N’Dama cattle. Bertalanffy’s model had the best Asymptotic estimate (A) for males’ data and could predict 99.40 per cent of their mature weight, while Richards’ model, which had the best A’ estimate in females could predict 99.43 per cent of their mature weight. © GSP, India.