Exigências nutricionais de energia e predição do consumo de água em ovinos deslanados

Abstract

The aim of this study was to estimate the nutritional energy requirements for the maintenance and gain of Santa Inês x Dorper females and to develop mathematical models for prediction of water intake (WI) for hair sheep. In the first study, 28 Santa Inês x Dorper females were used in a completely randomized design with three dietary levels (ad libitum, 30 and 60% feed restriction). At the beginning of the experimental trial, four animals were slaughtered as a reference group to estimate the initial empty body weight (EBWi) and initial body composition of the remaining animals. The NEm was estimated at 77.31 kcal/kg0.75 EBW day and the km was 63.80%. The model to estimate NEg (Mcal/day) was: 0.397 (±0.068) × EBW0.75 × EBWG0.9168 (± 0.079) R2 = 0.99 RMSE=0.002. The kg was 35%. In conclusion, the energy requirements for maintaining Dorper x Santa Inês females are similar to those reported by International Committees NRC (2007) and CSIRO (2007). The energy requirements for gain are higher than those reported by International Committees NRC (2007) and CSIRO (2007). In the second study, the data set used was derived from eight studies containing 185 individual observations of hair sheep with three sex classes (120 non-castrated males, 22 castrated and 43 females) in a feedlot system. The stepwise procedure used an ordinary least squares regression with significance of P < 0.05 to determine which variables would be included in the prediction model. Then, a random coefficient model was used, considering the random effect of the study and including the possibility of covariance between the intercept and slope. The sex class was considered a fixed effect and tested in the model parameters. For the model validation procedure, the comparison between predicted and observed values was carried out using the Model Evaluation System software. The sex class did not affect (P = 0.3340) the model for predicting the WI of hair sheep, with a single equation generated: WI (kg/day) = 0.1282 (± 0.5861) + 2.4186 (± 0.5842) x (R2=0.70; MSE= 0.1631; AIC = 297.6). Validation indicated that the suggested model accurately predicts the CA of hair sheep (P > 0.05 for the intercept) and (β0= 0) and slope (β1= 1). In conclusion, the prediction model allows us to accurately predict WI in hair sheep.