Modelos para estimação de componentes de (co)variância para produção de leite no dia do controle de vacas da raça Holandesa
Carregando...
Data
2010-02-26
Autores
Bignardi, Annaiza Braga [UNESP]
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
Parâmetros genéticos para a produção de leite no dia do controle (PLDC) de primeiras lactações de vacas da raça Holandesa foram estimados utilizando os modelos multicaracterísticas e modelos de regressão aleatória (MRA). Para os modelos multicaracterísticas foram analisados 15.896 controles mensais de produção de leite de 1.820 primeiras lactações de vacas da raça Holandesa. As análises foram realizadas por meio de sete modelos: multicaracterísticas padrão, três modelos de posto reduzido ajustando os primeiros 2,3 e 4 componentes principais genéticos e três modelos utilizando análise de fatores com 2,3 e 4 fatores. Para todos os modelos foram considerados os efeitos aleatórios genético aditivo e residual, e os efeitos sistemáticos do grupo de contemporâneo e da idade da vaca ao parto (efeito linear e quadrática) e do número de dias em lactação (efeito linear). A matriz de (co)variâncias residual, para todos os modelos, foi assumida ter posto completo. Os resultados indicam que somente dois componentes principais são requeridos para modelar a estrutura de (co)variâncias genéticas entre as produções de leite no dia do controle. Além disso, o modelo de posto reduzido diminui consideravelmente o número de parâmetros, sem reduzir a qualidade de ajuste. Para os MRA foram analisados 152.145 controles semanais de produção de leite de 7.317 primeiras lactações de vacas da raça Holandesa, provenientes de rebanhos da região Sudeste do Brasil. As produções de leite no dia do controle (PLDC) foram consideradas em 44 classes semanais de dia em lactação. Os grupos de contemporâneos foram definidos como rebanho-ano-semana do controle compondo 2.539 classes e, contendo, no mínimo, seis animais. O modelo utilizado incluiu os efeitos aleatórios genético aditivo direto, de ambiente permanente e o residual. Foram considerados como efeitos fixos, o grupo de...
Genetic parameters for first lactation test-day milk yields of Holstein cattle were estimated using multivariate and random regression models (RRM). For multivariate model a total of 15,896 individual monthly test-day milk yields (10 test-days), from 1,820 complete first lactations of Holstein cattle. A standard multivariate analysis, reduced rank analyses fitting the first 2, 3 and 4 genetic principal components, and analyses that fitted a factor analytic structure considering 2, 3 and 4 factors, were carried out. All models included also fixed effects of the contemporary groups, age of cow (linear and quadratic effects) and days in milk (linear effect). The residual covariance matrix was assumed to have full rank throughout. The results indicate that only two principal components are required to model the genetic covariance structure among the test-days milk yield. Furthermore, reduced rank model allows decreasing the number of parameter without reducing the goodness of fit considerably. For RRM A total of 152,145 weekly test-day milk yield records from 7,317 first lactations of Holstein cows distributed across 93 herds in southeastern Brazil were analyzed. Test-day milk yields were classified into 44 weekly classes of days in milk (DIM). The contemporary groups were defined as herd-year-week of test-day. The model included direct additive genetic, permanent environmental and residual effects as random and fixed effects of contemporary group and age of cow at calving as covariable. Mean trends were modeled by a cubic regression on orthogonal polynomials of DIM, and additive genetic and permanent environmental effects were estimated using B-splines. Residual variances were modeled by step function with 6 variance classes. Although all the model selection criteria utilized indicated the model employing cubic B-splines to both random effects, with 6 knots, a more parsimonious model ... (Complete abstract click electronic access below)
Genetic parameters for first lactation test-day milk yields of Holstein cattle were estimated using multivariate and random regression models (RRM). For multivariate model a total of 15,896 individual monthly test-day milk yields (10 test-days), from 1,820 complete first lactations of Holstein cattle. A standard multivariate analysis, reduced rank analyses fitting the first 2, 3 and 4 genetic principal components, and analyses that fitted a factor analytic structure considering 2, 3 and 4 factors, were carried out. All models included also fixed effects of the contemporary groups, age of cow (linear and quadratic effects) and days in milk (linear effect). The residual covariance matrix was assumed to have full rank throughout. The results indicate that only two principal components are required to model the genetic covariance structure among the test-days milk yield. Furthermore, reduced rank model allows decreasing the number of parameter without reducing the goodness of fit considerably. For RRM A total of 152,145 weekly test-day milk yield records from 7,317 first lactations of Holstein cows distributed across 93 herds in southeastern Brazil were analyzed. Test-day milk yields were classified into 44 weekly classes of days in milk (DIM). The contemporary groups were defined as herd-year-week of test-day. The model included direct additive genetic, permanent environmental and residual effects as random and fixed effects of contemporary group and age of cow at calving as covariable. Mean trends were modeled by a cubic regression on orthogonal polynomials of DIM, and additive genetic and permanent environmental effects were estimated using B-splines. Residual variances were modeled by step function with 6 variance classes. Although all the model selection criteria utilized indicated the model employing cubic B-splines to both random effects, with 6 knots, a more parsimonious model ... (Complete abstract click electronic access below)
Descrição
Palavras-chave
Leite - Produção, Funções de covariância, Modelos reduzidos, Parâmetros genéticos, Covariance functions, Genetic parameter, Reduced rank model, Splines, Test-day
Como citar
BIGNARDI, Annaiza Braga. Modelos para estimação de componentes de (co)variância para produção de leite no dia do controle de vacas da raça Holandesa. 2010. xiii, 71 f. Tese (doutorado) - Universidade Estadual Paulista, Faculdade de Ciências Agrárias e Veterinárias, 2010.