Ajuste de modelos de crecimiento y estudio del efecto de la oferta de forraje y del grupo genético de las vacas sobre la evolución de peso vivo de sus hijas hasta los 26 meses de edad

Since 2007, an experiment has been carried out at the Professor Bernardo Rosengurtt Experimental Station to evaluate the effect of modifying the supply of natural pasture forage on the productive and reproductive performance of breeding cows of various genetic groups. A factorial design was used that combined two levels of pre- and postpartum forage supply: High and Low, with two levels of the genetic group factor: Purebred and Crossbred, resulting in the following treatments: High Crossbred, High Purebred, Low Crossbred and Low Purebred. The objective of this particular work is to evaluate the effect of the forage supply with which the cows were managed during the pre- and postpartum and their genetic group on the live weight of their daughters from birth to 26 months of age. A database consisting of 2373 live weight records was used, corresponding to 182 heifers, daughters of the cows managed in the aforementioned experiment, which were born between 2007 and 2014. Considering the live weight of the heifers as the variable of interest, a mixed model was formulated in which the forage supply at which the heifer's mother was managed, the genetic group of the mother, the age of the heifer, as well as the double interactions between the age of the heifers and the other two factors mentioned above were included as fixed effects. The heifer and the year of its birth were considered as random effects. Based on this model, the forage supply assigned to the mothers of the heifers, as well as the interactions between age and supply and between age and the genetic group of the mother, had a significant effect on the live weight of the heifers, while the genetic group of the mother did not. Taking into account the factors included as fixed effects in the model, Tukey tests were performed to analyze each of them. The results of these tests showed no significant differences between the weights of the heifers at each of the ages evaluated considering, on the one hand, the levels of forage supply (high vs. low) and, on the other hand, the levels of the genetic group factor (purebred vs. crossbred). Regarding the individual effects analyzed, significant differences were detected between the levels of forage supply, with heifers from cows managed at a high forage supply being heavier than those from cows managed at a low supply (208 vs. 205 ± 5 kg). On the other hand, no significant differences were found in the average live weight of the heifers according to the genetic group of their mothers. Additionally, another mixed model was proposed with the aim of evaluating a series of statistical models that described the growth experienced by the heifers during the analysis period. This model considered the live weight of the heifers as the variable of interest and, as fixed effects, the treatment received by the mothers, the genotype of the father and the age of the heifers expressed in days. The heifer and the year of birth were included as random effects. Linear regression models (such as the first and second order models) and non-linear regression models (such as the Brody, Von Bertalanffy, Gompertz and Logistic models) were evaluated. All models showed a very high fit to the data, with adjusted r2 values ​​around 0.93, with practically no differences regarding this coefficient between the evaluated models. Regarding the AIC and BIC coefficients achieved by these models, it was found that, within the linear regression models, the second order model was the one that presented the lowest values ​​of said coefficients, while, within the non-linear regression models, it was the Brody model. Therefore, within the group of linear regression models, the second-order model was the model that best fit the database, while within the group of nonlinear regression models, it was the Brody model.