Study On Nonlinear Mixed Effect Model For Evaluation Of Muscovy Body Weight For Growth

In this study, 116 Muscovy ducks were fed and observed for 21 weeks at days 0, 7, 14, 21, 28, 35, 42, 49, 56, 63, 70, 77, 84, 91, 98, 105, 112, 119, 126, 133, 140, 147 to describe and estimate the trajectory and parameter for body weight growth of native Muscovy, and investigate and make clear the regular pattern that the body weight growth of native Muscovy varies with time. Five body weight growth functions, Gompertz, Logistic, Von Bertalanffy, Richards, and Brody, were estimated using hybrid genetic algorithm. Nonlinear fixed and mixed effect models were fitted to data of body weight in Muscovy The 8 growth parameters, mature weight, Point of inflection (POI), POI weight, birth weight, absolute growth rate, relative growth rate, absolute maturing rate, and relative maturing rate, were computed for these models. The results of fitting and estimating with the different growth models were compared using residual error variance, correlation index, and information criteria for evaluating the alternative models. The curves of body weight growth for every bird were plotted, and individual BLUPs of mature body weight of the Muscovy ducks were simulated by Bayesian-MCMC.The results from observing and analyzing showed: 1) there are obvious four phases during growing in the body weight of Muscovy ducks. Earlier stage of their growing is d 0 to 21. Average daily gain (ADG) of male and female Muscovy ducks was 15.7 g Id. Their body weight is obviously different between male and female birds after 21 of age. Quickly growing stage in body weight of them is d 21 to 49 for male birds, d 21 to 42 for female birds. ADG was 55.2 g/d in males, 41.9 g /d in females. Slowly growing stage in body weight of them is d 49 to 70 for male birds, d 49 to 70 for female birds. ADG was 40.5 g/d in males, 27.0 g/d in females. Later stage of their growing is d 77 to 147 for males, d 70 to 147 for females. Average daily gain was 7.3 g /d in males, 3.2 g /d in females. Their body weights were nearly stable during this phase. 2) The 5 nonlinear fixed effect models with 8 growth parameters each were fit for body weight of Muscovy ducks. Residual error variance of Richards growth model was least and had a 18.89% reduction, correlation index totally maximum compared with these of the other 4 growth models. The 8 growth parameters estimated by these nonlinear fixed effect models had larger accuracy compared with the actual observed data as a whole. 3) The 5 nonlinear mixed effect models with 8 growth parameters each were fit for body weight of Muscovy ducks. Their accuracies had a much larger increase than these of the nonlinear fixed effect models. Among the 5 nonlinear mixed effect models, Information criteria of Gompertz growth model were least, and Reduction of residual error variance of this model took the second place, up to 99.82%. It is a best growth model for fitting the body growth for Muscovy group. For the body growth for Male or female Muscovy group, Information criteria of Richards growth model was least among the 5 nonlinear mixed effect models. Reduction of residual error variance of this model took the first place, up to 89.80% for male group, 91.81% for female group. It is a best growth model for fitting the body growth for Muscovy male or female group. By means of Bayesian-MCMC simulation, phenotypic values of BLUP of individual mature weights predicted by the best nonlinear mixed effect models for fitting the body weight of Muscovy, Gompertz and Richards, were changed between observed and posterior samples, BLUPs of some individuals were changed from the. positive to the negative, and others had non-significant differences by testing. Thus these are propitious to evaluation of breeding values for individuals.Based on the body growth characteristic of Muscovy duck, Richards nonlinear mixed effect model describing equation and parameter for its body growth at different ages, and predicting BLUP breeding value for an individual is accurate than the nonlinear fixed effect model, can be recommended to apply to breed selecting, ducklings producing, rearing management and economic utilization to obtain some information of valuing.