Avaliação de estratégias de seleção em programa de melhoramento genético de suínos por meio de simulação de dados

Abstract

The objective of this study was to evaluate, through simulation of data, different selection strategies adopted in a breeding program of swine and their impacts on levels of inbreeding and genetic gains. In addition to specifically estimating the impact of restrictions on the maximum number of sons and daughters by male or female and the impact of rates and productive losses on levels of inbreeding and genetic gain. Data from real populations A, composed of Pietrain pigs, and B, by Landrace pigs, used in this study, came from two genetic lines located on farms in the west of the Santa Catarina state. To generate the simulated populations, a Fortran language simulator was used in which the information of the real populations was used: two main input files, one containing the pedigree of the last 10 years, with 21,906 animals in population A and 251,343 animals in population B, And another one containing the estimated breeding values for age, backfat and feed conversion, all adjusted to 110 kg of live weight, for both populations, as well as depth of the longissimus dorsi muscle adjusted for 110 kg of live weight - only for population A, and number of live piglets at the 5th day of life, per farrowing, only for population B, of the selected animals in 2014 (Generation 0). In addition to the (co)variances of estimated breeding values, rates and productive and reproductive averages, number of mating and number of animals selected per generation. In each article, three scenarios were simulated: in article 1 the scenarios varied in the restrictions on the number of full siblings and half-siblings selected, for males and females, already in article 2 the variations in the scenarios were in the mortality rate in the lactation and farrowing rate. Ten generations were simulated, with 30 repetitions each generation and scenario. The results of the simulation of data in a breeding program of swine allow to conclude that: there is an increase in the inbreeding levels in a closed nucleus independent of the selection strategy used; the rate of inbreeding are larger in populations of smaller effective size; restrictions on the number of full siblings and half-siblings selected are efficient to reduce the rates of inbreeding, with the restriction of a maximum of two full-siblings and three half-siblings for males and three full-siblings for females, for having obtained the highest genetic gains, is indicated as a selection strategy to be adopted in these populations; there is an increase in the levels of consanguinity in a closed nucleus with an increase in productive and reproductive losses; the productive and reproductive losses reduce the variances of the estimated breeding values and, mainly, the intensities of selection, reducing the genetic gains; actions that maximize farrowing rates are preponderant to those that minimize mortality rates in the lactation, since the reduction in simulated farrowing rate has resulted in greater losses in genetic gains.