Parametrização e validação de um modelo de estimativas das exigências e do desempenho de gado de corte / Parametrization and validation of a model to estimate the requirements and performance of beef cattle

AUTOR(ES)

Beatriz Motta Hoffmann

DATA DE PUBLICAÇÃO

2007

RESUMO

Although simulation models to estimate the nutritional requirements and the performance of beef cattle exist (e.g. National Research Council – NRC, 1996), these have been developed mostly for temperate conditions and European cattle. There are no update publications with information on nutritional requirements and feeding recommendations for ruminants. The objective of this study was to parametrize and then validate a model based on the California net energy system, and incorporate recent results of research with Bos indicus (mostly Nellore) cattle and tropical feeds. After the inclusion of new equations (e.g. for feed intake) and new equation parameters, the resulting model was evaluated in two manners: first using independent experimental results and, later using results obtained in commercial feedlots of Brazil (from the central region). The accuracy and the precision of the model were tested getting the model predicted dry matter intake (DMI) and shrunk daily weight gain (SDWG) and comparing these with observed values. Linear regressions between predicted and observed values and the residues of predicted less observed were plotted and described as a function of certain variables. The model predicted DMI and SDWG in experimental situations with R2 of 67,9 and 48,7%, respectively. For the regression analysis of DMI from commercial feedlots results the R2 were of 85,0 and 74,0% for crossbred and Nellore bulls, respectively. However for steers the model predictions were less efficient with R2 of 62,4% for crossbred. The lower efficiency of these predictions seems to be associated to large variability in frame score and the intensity of the compensatory growth, since these variables were not adequately described in our data. A better description of the animals may improve the potential of the model to estimate gain and feed intake. Additionally, the orts were not collected, and this may suggest that the observed DMI were not in fact the real DMI. In the prediction of gains in commercial feedlots, the model tended to overestimate probably because these predictions were calculated from the observed DMI, which do not incorporate the orts. The model did not estimate well the long-term performance of animals which grew slower and had to be fed longer. These animals tend to gain less weight and deposit more fat. This can be explained because compensatory growth is a short-term event, and animals fed longer do not express compensatory growth during the second half of the feeding period. This suggests that the simulation must be made in two stages. The lower efficiency of the predictions for commercial data seems to be related to the imperfections in the description of compensatory growth, in the evaluation of frame size and in not accounting of the orts on the feeders.

ASSUNTO(S)

weight gain bovino de corte confinamento animal requirements nutrição animal intake metabolismo animal model modelos matemáticos

Documentos Relacionados

• Desempenho e exigências de energia, proteína e minerais de bovinos de corte em pastejo, submetidos a diferentes estratégias de suplementação
• Desenvolvimento de um modelo para estimativas da produção de gases de efeito estufa em diferentes sistemas de produção de bovinos de corte
• Aplicação de modelo de gestão por macroprocessos e sistema de indicadores de desempenho em empresa de pecuária de corte
• Um modelo baseado em conhecimento para simular rebanhos de bovinos de corte
• Desempenho produtivo e exigências nutricionais de bovinos de corte em pastagens de Brachiaria decumbens, com suplementação protéica