Application of different growth models to “Nero di Parma” pigs

Abstract

The growth curves of 280 “Nero di Parma” pigs from birth to maturity were calculated by applying nine different models (regressions from 1st to 4th degree and nonlinear regressions following the Brody, Logistic, Janoschek, Bertalannfy and Gompertz models) to 1109 individual records of body weight (BW) from two different data sets. The goodness of fit of experimental data was calculated by means of Residual Variance, Akaike Information Criterion, Residual Standard Deviation and R2. The best fit was obtained by Gompertz equation, as follows: BW(kg)= 240.2±2.4 * esp (-exp (-0.0069±0.0001*(age(d)–213.5±3.1))). Regardless to the model, all correlations between actual and estimated BW were highly significant (P<0.001): the highest correlation (0.980) was obtained by the application of the Gompertz equation. In conclusion the growth of “Nero di Parma” pigs can be well described by applying the Gompertz model to field data.