Modelos não lineares em caprinocultura leiteira e análise econômica do sistema de produção

• Ano de defesa: 2016
• Autor(a) principal: Leonardo de Rago Nery Alves
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais; UFMG
• Palavras-chave em Português: Economia; Modelagem; Cabras; Produção; Leite; Leite Produção; Produção animal; Caprino Aspectos econômicos; Lactação
• Link de acesso: http://hdl.handle.net/1843/BUOS-B6FDYR

Mathematical modeling is a feature that aims to contribute to the knowledge linked to animal production, and allow quantification of several variables in animal perspective. This study aimed to characterize the biometric growth in goats, as well as adjust the lactation curve of Saanen goats and study the economic feasibility of the production system. The experiment took place in a property located in the municipality Bonfim (Minas Gerais). The study of biometric growth, were used 44 goats, accompanied by the birth to the clinical manifestation of the puberty, totaling 248 days. For lactation curves were used 1,167 average daily milk production records (kg) produced by 89 Saanen goats that produced 168 lactation profiles. The profiles were analyzed according to the order of lactation (OL) and type of birth (TN). The effect of OL was observed until the fifth lactation (OL1, OL2, OL3, OL4 e OL5). The number of records and lactation profiles for each OL were : OL1 = 458;56, OL2 = 208;30, OL3 = 116;14, OL4 = 151;19 e OL5 = 147;17. For TN were checked the effect of single births (TN1), double (TN2) and triple (TN3). We used different non- linear models and variance functions. In the economic study we used financial information of the property during the years 2012-2015, and tested the feasibility indicators were Gross margin, net margin and profit. In biometric evaluation, the models were: Gompertz for body mass and heart girth, Brody for height withers, croup height and body length; Linear to croup width and Logistic for length of croup. The Scaled variance function explained the behavior of the variance of all variables except for croup width that showed exponential response. Wood model and Scaled variance function was the most suitable to estimate the effect of various lactation orders on the curve and the exponential decay with staggered variance function was the most likely to type of birth effect on the curve parameters. The system showed average profit R$ 14,255.95 for the evaluated years, and economical viable.