COMPARATIVE ANALYSIS OF THREE GROWTH MODELS IN INDIGENOUS NORMAL FEATHERED CHICKENS OF NIGERIA

This study was carried out to compare three growth models using body weights of 123 progenies generated from unselected random bred parents. The experiment which lasted for 24 weeks was carried out at the Poultry production unit (PPU) of the Ministry of Animal and Fisheries Resources, Potiskum, Yobe state. The data were analysed using Statistix (Version 9.0). Coefficient of determination (R2) values for Gompertz, Logistic and Richard models were 0.61, 0.73 and 0.44 respectively. The Logistic model had the highest R2 value (0.73) and Richard had the least (0.44). In contrast, Richard had the highest MSE (Mean Square Error) and AIC (Akaike’s Information Criterion) while, Logistic had the least. Richard had the poorest fit (high MSE, AIC and lower R2) while Logistic model and Gompertz best described growth in that order (lower MSE, AIC and high R2). Consequently, based on the goodness of fit criteria; R2, MSE and AIC values, the logistic function best described the growth pattern in indigenous normal feathered chickens of Nigeria.