Application of growth models to different sheep breed types in South Africa

Abstract

Abstract In order to determine the growth trends of seven sheep breeds that are popular in South African commercial production systems (Dohne Merino, Dormer, Dorper, Meatmaster, Merino, SA Mutton Merino and White Dorper), body weight data was collected from lambs of each breed from birth until they were assumed to have attained maturity at one year of age. These breeds were selected to represent wool, meat and dual purpose types, as well as a range of maturity types, ranging from early to late maturing. From the data, individual growth curves were extrapolated and the Brody, Gompertz, Logistic and Von Bertalanffy models fitted to the curves. The model parameters were compared between breeds for each model fitted and the respective models evaluated for describing the growth in various breeds. The Brody model was found to be unsuitable, as the model over-predicted the mature weights of the sheep and presented the highest RMSE and AIC values, as well as the lowest R2 values for each breed. The Gompertz, Logistic and Von Bertalanffy models were found to be more suitable in accurately predicting the growth of the different production groups. The Logistic model presented the lowest RMSE and AIC values, and highest R2 values, in each case where the model was fitted to the production group and is the most accurate in predicting the growth of the slaughter lambs of the breeds in this study.