Avaliação genética e estudo do crescimento de matrizes de codorna de corte utilizando modelos de regressão aleatória

• Autor(a) principal: Teixeira, Rafael Bastos
• Data de Publicação: 2009
• Tipo de documento: Tese
• Idioma: por
• Título da fonte: LOCUS Repositório Institucional da UFV
• Texto Completo: http://locus.ufv.br/handle/123456789/1853
• topic: Avaliação genética; Simulação de dados; Controle de acasalamentos; Endogamia tamanho efetivo; Genetic evaluation; Simulation of data; Control of matings; Inbreeding effective size; CNPQ::CIENCIAS AGRARIAS::ZOOTECNIA::GENETICA E MELHORAMENTO DOS ANIMAIS DOMESTICOS

Data from two genetic groups of meat-type quail hens were evaluated, first group with records from 1.026 animals from genetic group UFV1, and the second, with records from 1.110 animal from UFV2. The evaluated traits were: Live weight (P0, P7, P14, P21, P28, P35, P42, P77, P112 e P147), average egg weight (POM1, POM2, POM3 e POM4), average weight of egg shell (PCM1, PCM2, PCM3 e PCM4), average yolk weight (PGM1, PGM2, PGM3 e PGM4), average albumen weight (PAM1, PAM2, PAM3 e PAM4), average specific gravity of the egg (DM1, DM2, DM3 e DM4), average width of the egg (LOM1, LOM2, LOM3 e LOM4), average length of the egg (COM1, COM2, COM3 e COM4), number off eggs (N1, N2, N3 e N4) and age of first egg (IDPO). Statistical variance and test means (Fischer at 5%) were carried out and single trait analysis were made to estimate components of variance. Principal component analysis were made to define variables of major importance on the studied populations. Two-trait analysis were also carried out for estimation of covariance and genetic correlations of the selected traits from principal component technique. UFV2 group was superior in all traits. Heritabilities from UFV1 were POM1(0,27), POM2(0,30), POM3(0,54), POM4(0,43), PGM1(0,13), PGM2(0,21), PGM3(0,21), PGM4(0,26), PCM1(0,41), PCM2(0,46), PCM3(0,38), PCM4(0,47), PAM1(0,38), PAM2(0,42), PAM3(0,54), PAM4(0,54), DM1(0,31), DM2(0,15), DM3(0,36), DM4(0,31), IDPO(0,16) and TXT(0,10). Heritabilities of the selected traits on UFV2 were POM1(0,17), POM2(0,36), POM3(0,55), POM4(0,40), PGM1(0,18), PGM2(0,30), PGM3(0,37), PGM4(0,33), PCM1(0,21), PCM2(0,33), PCM3(0,46), PCM4(0,49), PAM1(0,25), PAM2(0,29), PAM3(0,55), PAM4(0,35), DM1(0,18), DM2(0,16), DM3(0,44), DM4(0,22), IDPO(0,10) and TXT(0,03). Principal components analysis pointed variables P7, P14, P77, P147, LOM4, COM2, PGM1, PGM2 and IDPO as main important for UFV1. From 31 recorded traits, 9 were responsible for 81.85% of total variation. For UFV2, P7, P35, P147, COM1, COM2, COM3, PCM3, PAM4 and TXT were pointed as main important, and 9, out of 30 recorded traits, were responsible for 79.91% of total variation. Two possible models of heterogenic residual were tested being grouped in three age classes: CL3: 1-7-14, 21-28-35 days and 42-77-112-147 days and five age classes: CL5: 1-7, 14-21, 28-35, 42-77 e 112-147 days of age. After choosing the residual variance to be used on the analysis, an evaluation was carried out to find out the best random regression model to fit the growth curve of the hens. The comparison among the models were done by Akaike s information Criteria (AIC), Schwartz s Bayesian information Criteria (BIC), logarithm likelihood function (Loge L) and Likelihood ratio test, 1%. The model considering heterogeneity residual variance showed to be reliable for UFV1 group and model with 5 classes for UFV2. The use of Legendre s polynomial, with 5 orders, for direct genetic additive effect and 5, for permanent animal effect for UFV1 and orders 3, for for direct genetic additive effect and 5, for permanent animal effect for UFV2, should be used for the evaluation of growth curve for the studied meattype quail hens.