Evaluation of growth functions on Japanese quail lines

Abstract

The aim of present studywas to fit the best predictive equation to describe the growth curve of different Japanese quail lines. Moreover the effect of short-term divergent selection on the growth curve parameters was investigated. The quail lines utilized in the current studywere two divergentlyselected lines for high (HW) and low (LW) 4-wk bodyweight (BW) and a control line (C). Determination and adjusted determination coefficients, relative error mean and standard deviation, mean square error, Akaike's information criteria and Schwarz Bayesian information criteria were used to evaluate the accuracyof prediction with the growth functions of Hyperbolastic (H1, H2, H3), Richards, Gompertz, Logistic and Von bertalanffy. Based on model behavior and statistical performance, the Gompertz and Logistic functions were not able to show a suitable fit for all three lines. The overall goodness of fit statistics in the HW line showed that the Richards function has the best fit to the data followed byH3, H2, H1, Von bertalanffy, Gompertz and Logistic functions, respectively. The overall results in the LWand C lines were similar to the HW line, except that Logistic function provided a better fit to the data than Gompertz. The studyof growth pattern using Richards function revealed that short-term divergent selection altered the growth trajectoryof selected lines through the changing of shape parameters and relative intensityof growth rates. © 2013, Japan Poultry Science Association.