Nonlinear modeling of growth and study of a new characterization point applied to Timahdit breed sheep under uncertain paternity

Abstract

The objectives of this study are to determine which model, among those of Brody, Logistic, Gompertz, Von Bertalanffy and Richards, is the most appropriate to describe the growth of this breed and to choose a point for this model which would be closest to the slaughter weight which have to be used as a selection criterion. Also, in preparation for genetic evaluation under uncertain paternity, Henderson's Average Numerator Relationship Matrix method was used and an R code was developed for the creation of the inverse matrix A which was replaced the pedigree in the animal model. To do this, 64,282 observations corresponding to 12,944 animals, collected between 2009 and 2016, were analyzed. The Von Bertalanffy function had the smallest AIC, BIC, and deviance criteria values, indicating better data representation for both sexes. Considering the average slaughter live weight of 29.4 kg in the study area, the new point of characterization named f(t_bm), later than the inflection point of the growth curve, is closer to the commercial target for females intended for regular slaughter supplying butchers and for animals of both sexes intended for religious festivals. Therefore, it would be appropriate to consider this point as a selection criterion for this breed. The developed R code will be integrated into a free R package allowing the estimation of the genetic parameters of the traits of the Von Bertalanffy model.

Keywords: Body weight; Growth model; Nonlinear regression; Timahdit breed; Uncertain paternity.