Use este identificador para citar ou linkar para este item: http://repositorio.ufla.br/jspui/handle/1/59772
Título: Modelos para descrever a curva de lactação de ovelhas Santa Inês e mestiças Santa Inês x Lacaune
Título(s) alternativo(s): Models to describe the lactation curve of Santa Inês sheep and Santa Inês x Lacaune Crossbred Sheep
Autores: Muniz, Joel Augusto
Fernandes, Tales Jesus
Savian, Taciana Villela
Palavras-chave: Modelos não lineares
Modelo Wood
Ovinocultura leiteira
Nonlinear models
Wood model
Dairy sheep farming
Data do documento: 20-Dez-2024
Editor: Universidade Federal de Lavras
Citação: CARVALHO. Louiziane Ribeiro. Modelos para descrever a curva de lactação de ovelhas Santa Inês e mestiças Santa Inês x Lacaune. 2024. 69 p. Dissertação (Mestrado em Estatística e Experimentação Agropecuária) - Universidade Federal de Lavras, 2024.
Resumo: Dairy sheep farming, although a smaller segment compared to sheep meat production, plays a significant role in various regions of the world. Sheep milk is more concentrated than goat and cow milk, serving as the basis for high-value-added dairy products. Sheep milk production in Brazil is relatively recent, with the Lacaune breed, a French dairy breed, standing out. In contrast, the Santa Inês breed, of Brazilian origin, is specialized in meat production. The lactation curve, which represents milk production over time, is an important tool for nutritional management and genetic improvement of the herd. The objective of this study was to compare different regression models to describe the lactation curve of Santa Inês sheep and crossbred sheep (a mix of Santa Inês and Lacaune breeds). Additionally, equations were derived to estimate the lactation peak and total milk production for the studied models. The data analyzed were extracted from the study by Ferreira (2009). The models proposed by Brody et al. (1924), Nelder (1966), Wood (1967), Cappio-Borlino, Pulina, and Rossi (1995), and Cobuci et al. (2000) were used. Model parameters were estimated using the least squares method with the Gauss-Newton convergence algorithm in R software. Residual analysis included normality (Shapiro-Wilk), homoscedasticity (Breusch-Pagan), and independence (Durbin-Watson) tests. Model fit quality was assessed using the adjusted coefficient of determination (Raj2) and the Akaike Information Criterion (AIC). Based on the evaluation criteria, fit quality, practical interpretability of model parameters, and the ability to estimate lactation persistence, the Wood model was considered the best for describing the lactation curve of sheep. The parameter estimates for the Wood model were: a= 1.033, b = 0.267, and c = 0.061 for Santa Inês sheep, and a = 1.611, b = 0.434, and c = 0.102 for crossbred sheep. The lactation peak, as estimated by the Wood model, was 1.174 liters for Santa Inês sheep and 1.954 liters for crossbred sheep, both occurring between the 4th and 5th weeks of lactation. Total milk production was 108 liters for Santa Inês sheep and 164 liters for crossbred sheep. Lactation persistence, estimated by the Wood model, was 3.55 for Santa Inês and 3.27 for crossbred sheep. Equations and estimates were derived to characterize the lactation peak and total milk production for the models under study.
URI: http://repositorio.ufla.br/jspui/handle/1/59772
Aparece nas coleções:Estatística e Experimentação Agropecuária - Mestrado (Dissertações)



Este item está licenciada sob uma Licença Creative Commons Creative Commons