Comparison of Linear with Some non Linear Regression Methods to Estimate Hot Carcass Weight Using Live Weight in Arsi-Bale Sheep and Goats of Both the Sexes

The study was conducted on estimation of eviscerated carcass weight using live weight as predictor for Arsi Bale sheep and goat of both the sexes, reared under farmer’s management and on natural pasture. The age group of the studied sheep and goat were <1 year (category 1) and >1 but < 2 years (category 2). The age of the animals were estimated by dentition method. The results indicated that the average slaughter weight of the studied animals were lower than those reported, this may be ascribed to poor nutrition obtained from the pasture grasses and also due to presence of internal parasites. The dressing percentage as assessed in the study was higher in the goats and among the goats was higher in the bucks. The study also indicates that the among the non linear and linear regression equations the best predictor for the carcass weight in rams in category 1 was by using Quadratic regression equation, while for the age category 2 it was by using the Sigmoid regression equation. The results for prediction of carcass weight in ewes indicate that the Quadratic regression equation provides the highest R values for both the age categories. The regression results also 2 shows that the best predictor for estimation of carcass weight in bucks for the age category 1 was by using the Quadratic regression method, while that for age category 2 was by using Logarithmic, growth, exponential and logistic regression method ( the R values being same for all). The results related to estimation of carcass weight 2 of the does indicate that the best predictor for age category 1 was by taking into account the Quadratic regression equation and that for age category 2 was by considering the power regression equation.