Modeling growth of Nigerian indigenous and tropically adapted chicken genotypes using developmental parameters

Abstract

The possibility of modeling growth with the aim of visualizing growth patterns over time, and generating equations that can be used to predict the expected weight of the animal at specific age could be an impetus for optimization of farmer’s livelihood. The weekly body weight of 993 off-springs of seven genotypes of chicken, consisting of Nera Black-NB, White Leghorn-WL, Giriraja-GR, Naked Neck-NN, Frizzle Feather-FF, Normal Feather-NF and FUNAAB Alpha chicken-BA, were fitted to Logistic, Gompertz, Richards and Bertalanffy growth model using the procedure of NLIN (Marquart algorithm) based on Restricted Maximum Likelihood approach (ReML). The study revealed that GR chickens performed better than other genotypes, while BA had superior performance compared to the indigenous and the WL chickens. However, among the indigenous, the performance of NN chickens was best. There was a negative correlated relationship observed between asymptotic weight (A) and maturing rate (k). Gompertz model best fit the chicken data according to Akaike Information Criteria (AIC) and Bayesian Information Criteria (BIC) for FF, NF, and GR while Richards model on the other hand, had better fit for NN, NB and WL. Bertalanffy model was the best for BA chicken. The study concludes that high k will produce smaller A. Furthermore, mixing of improved exotic genes with the indigenous produces improved and better adapted genotypes in BA. AIC and BIC with ReML approach presented Gompertz model with wide applicability among the indigenous chickens while Richards model fit well for the locally adapted exotic chickens.