Prediction of live body weight from linear body measurements of west African long-legged and west African dwarf sheep in northern Ghana.

The knowledge of live weight of animals is so important in the livestock production and marketing practices that this study was undertaken to develop models for predicting the weight of sheep at market ages. Data comprising of the weight and linear body measurements were collected on the West African Long-Legged (WALL) and the West African Dwarf (WAD) sheep from PongTamale and subjected to regression analyses. The results revealed that heart girth was the best predictor of liveweight, with prediction accuracies of 92.36% for two years old WALL sheep and 81.20% for one year old WAD sheep, while wither height was the second most important trait in liveweight prediction, in simple linear models. The quadratic models of the single-trait models also had heart girth as the best predictor of liveweight, recording 92.92% accuracy for one year old WALL sheep. Only two traits were mostly required for weight estimation in the multiple-trait models, and the best model was obtained from two years old WALL where heart girth and body length accounted for about 95.53% in prediction accuracy. The multiple-trait quadratic models were generally better in liveweight prediction compared to the respective linear models. Clearly, weight estimation was more accurate among the WALL than the WAD sheep, and also among the younger sheep regardless of the breed. The variations in the models suggest that breed and age of sheep had influence on the type of models required to predict their live body weight.