Avaliação genética de suínos utilizando abordagens freqüentistas e bayesianas

• Autor(a) principal: Barbosa, Leandro
• Data de Publicação: 2007
• Tipo de documento: Tese
• Idioma: por
• Título da fonte: LOCUS Repositório Institucional da UFV
• Texto Completo: http://locus.ufv.br/handle/123456789/1419
• topic: Parâmetros genéticos; Máxima verossimilhança restrita; Amostrados de Gibbs; Genetic parameters; Restricted maximum likelihood; Gibbs sampling; CNPQ::CIENCIAS AGRARIAS::ZOOTECNIA::GENETICA E MELHORAMENTO DOS ANIMAIS DOMESTICOS

This study aimed to estimate (co)variance components and genetic parameters for economically important traits of a Large White swine population. The following traits were evaluated: days to 100 kg (DAYS) and backfat thickness adjusted to 100 kg (BF) as performance traits, and total number of born piglets (TNPB) as reproductive trait. (Co)variance components and genetic parameters were estimated by restricted maximum likelihood, with a derivative-free algorithm using the DFREML and MTDFREML software, and by Gibbs Sampling algorithm using MTGSAM software. This study was structured in four chapters, first and second for performance traits and third and fourth for reproductive traits. In the first chapter, four different mixed models were used and Akaike s information criterion (AIC) was used to choose the model that better fits to data. The inclusion of additive maternal genetic and common litter effects, in addition to the additive direct genetic effect, was recommended. The estimative of additive direct heritability were medium (around 0.28 and 0.45) and the maternal heritability were low (around 0.07 and 0.05) for DAYS and BT, respectively. Correlations between additive direct and maternal genetic effects were negative, showing an antagonism between these effects. The estimates of common litter effects were around 0.03 and 0.11 for BF and DAYS, respectively. In the second chapter, the (co)variance components and genetic parameters were estimated by Gibbs Sampling using the MTGSAM software. A mixed model with contemporary group was used as a fixed effect, while additive direct genetic, additive maternal genetic, common litter and residual as random effects. The additive direct heritabilities for DAYS and BF were 0.33 and 0.44, respectively. The mean estimates of common litter effects for DAYS and BF were 0.09 and 0.02, respectively. The estimate of additive genetic correlation between DAYS and BF was very close to zero (-0.015). In the third chapter, two mixed models (additive and repeatability model) were used to evaluate TNPB using the MTDFREML software. In the additive model, contemporary group was used as a fixed effect, while additive direct genetic and residual as random effects. In the repeatability model, parity order and permanent environment were included as fixed and random effect, respectively. The estimates of additive direct heritabilities were 0.15 and 0.20, for additive and repeatability models, respectively. The fraction of variance due to permanent environmental effects (c2) was 0.09. The estimates of genetic trends for additive and repeatability models were similar (around of 0.02 piglets/sow/year). In the fourth chapter, a mixed model with contemporary group was used as fixed effects, while additive direct genetic and residual as random effects. Data of the first four parities for TNPB were used in single trait and multitrait analyses, using TNPB in each parity as a different trait. Heritabilities of TNPB in different parities in single trait analyses ranged from 0.14 to 0.20. Estimates of additive direct heritabilities in multitrait analyses were similar to the estimates of single trait analyses. Estimates of phenotypic correlations were lower than genetics correlations. The estimates of genetics correlations were lower than 0.75 for all parities, except between third and fourth parities, which showed high genetic correlation (0.91). The smallest genetic correlation was observed between first and second parities (0.60).