Comparison of growth curve models in partridge
V Sariyel,A Aygun,I Keskin,V. Sariyel,A. Aygun,I. Keskin
DOI: https://doi.org/10.3382/ps/pew472
IF: 4.014
2017-06-01
Poultry Science
Abstract:This study was conducted in order to determine the goodness of fit of Brody, Gompertz, Logistic, and von Bertalanffy growth curve models in partridge (Alectoris chukar). The growth curve parameters A (upper asymptote or mature weight parameter), B (scale parameter related to initial weight), and K (instantaneous growth rate parameter) were determined as 623.4, 1.05, and 0.075 for females and 723.8, 1.06, and 0.073 for males in the Brody model, respectively, 472.9, 3.47, and 0.207 for females and 565.3, 3.59, and 0.192 for males in the Gompertz model, respectively, 440.2, 12.89, and 0.332 for females and 517.0, 14.13, and 0.319 for males in the Logistic model, respectively, and 498.9, 0.77, and 0.164 for females and 608.8, 0.79, and 0.150 for males in the von Bertalanffy model, respectively. While determining which growth curve model presented the better fit, the coefficient of determination (R2), adjusted the coefficient of determination (adj.R2), mean square predicted error (MSPE), Durbin-Watson value, correlation between estimated live weight and residual values (RESC), Akaike's information criteria (AIC), and Bayesian information criterion (BIC) were used. As a consequence of the study, it was determined that the Gompertz model yields a better fit to the data for the partridge, as its coefficient of determination and adjusted coefficient of determination are high, its values of MSPE, RESC, AIC, BIC are low and there is not an autocorrelation between the residual values. As a result, the Gompertz model presented a better fit with the partridge data.
agriculture, dairy & animal science