Prediction of internal egg quality traits of Potchefstroom Koekoek and Lohmann Brown layers using classification and regression tree method

Egg quality traits are features of eggs that influence the general quality of the egg. This study aimed to establish models for the prediction of albumen weight and yolk weight in Potchefstroom Koekoek and Lohmann Brown layers with a classification and regression tree (CART) decision methods. The Pearson's correlation findings displayed that the albumen weight had a positively high remarkable ( P < 0.01) association with the egg weight, egg width and yolk weight, while yolk weight had a positively high significant ( P < 0.01) relationship with the egg weight and albumen weight in Potchefstroom Koekoek. In Lohmann Brown, the albumen weight had a positively high remarkable ( P < 0.01) association with the egg weight, egg width and egg length, while yolk weight had a positively high significant ( P < 0.01) relationship with the egg weight, egg width and egg length. The CART method produced good models for predicting the albumen weight, with R 2 of 0.94 and 0.96, and yolk weight, with R 2 of 0.93 and 0.92, in Potchefstroom Koekoek and Lohmann Brown layers, respectively. The egg weight was shown to be the best leading predictor of albumen and yolk weight in both breeds. This study suggests that CART decision methods might assist in determining the breed standards of Lohmann Brown and Potchefstroom Koekoek chicken breeds in order for breeding programmes to improve their egg production. In conclusion, albumen weight and yolk weight can be improved best with enhancement of the egg weight.