Análisis de la fertilidad al primer servicio en novillas doble propósito bajo un modelo animal

in order of evaluating first service fertility (fps) on dual purpose heifers, this research was carry out with the electronic records from two commercial herds. a total of 6,823 records from crossbred animals. the response variable was first service fertility of the heifers. the analyses were realized under two approaches: classic and bayesian. for the classic analysis, it was used a simple animal model that included the discrete fixed effects of herd, breed group and year-season of service, as well as the covariates age at the first service and weaning weight, both in quadratic form. in addition, it included the random additive direct genetic effect of the animal and the residual. in the first case, the analyses were realized with the program mtdfreml. whereas in the second case, the data were analyzed with the same model animal in a threshold form, for which was used the program mtgsam. under this approach the size of the sampling chain, the burn in period and the thinning rate were approximated with gibbsit program, whereas the analysis of posterior distribution of the sampling chain were realized with program gibanal. the overall length of the chain was 136,000 samples, discarding the first 1,000 and settling down a thinning rate retaining each 14th element of the sequence. the global percentage of fps was of 65%. under the classic approach the heritability was 0.03 ± 0.02, which once transformed to the underlying normal scale raised to 0.05. meanwhile under the bayesian approach the analysis of the 9,642 samples reflected that the average of the posterior distribution of the heritability was 0.0735; the median of the same distribution was 0.0650, with estimators in the rank between 0.0130 and 0.2270. selection for this variable could only produce very slow changes in the population.