Comparison of nonlinear functions in male and female chickens at different seasons using restricted Maximum Likelihood Approach

Abstract

Weekly body weight data from 133 offsprings of Bovan Nera (NB), 93 of White Leghorn (WL), 105 of Giriraja (GR), 197 of Naked Neck (NN), 164 of Frizzle Feather (FF), 186 of Normal Feather (NF) and 115 of improved indigenous FUNAAB Alpha chicken (BA) were used for the study. The birds were individually wing tagged at day old for proper identification purpose. They were brooded in a brooding cage for the first three weeks then housed and reared on a deep litter until twenty weeks of age. Feed and water was given ad libtum. Weekly weight gain of the birds were subjected to four nonlinear growth models, Logistics, Gompertz, Richards and Bertalanffy to estimate the growth curve of the birds using fixed effect model and the procedure of NLIN (Marquart algorithm) The growth curves as predicted by the four models considered showed that both male and female GR followed by NB expressed better growth rate and highest weight at 20 weeks of age compared to other genotypes during late dry season and early wet season while the growth curve of WL in female and, WL and NF in male during late dry season had the lowest predicted curve. The predicted growth curve during early wet season had NF presenting the lowest growth curve for most of the model. The predicted growth curve showed a more consistence output in Logistic and Gompertz model. The goodness of fit favoured Gompertz as the model of best fit for predicting growth curve in both local and locally adapted chickens.