Identification of optimum test-day records for accurate prediction of lactation yield in Murrah buffaloes under farmers’ herd conditions

Accurate estimation of lactation yield under field conditions can increase the pace of selection and genetic improvement in Indian buffaloes. Monthly test day records can predict the 305-day milk yield. In the present study, daily lactation yield data from first and second lactation were used to identify the minimum number of test-day records, which could be used to predict the 305-day milk yield with accuracy and maximum correlation with actual lactation yield in Murrah buffaloes. Seventeen general linear models were developed using monthly test-day records and significant days of lactation. A combination of significant days were identified from the test-day model and the principal component analysis based significant days were used to build the final best fit model, which in the predicted data set, showed a comparatively lower RMSE (471.78) and higher R2 of 0.67. Our model showed a high correlationcoefficient of 0.82 with the test-day interval method of the International Council of Animal Recording for predicted lactation yield.