Multiple-trait prediction of lactation yields for dairy cows

Area of Science:

• Animal Science
• Dairy Science
• Quantitative Genetics

Background:

• Accurate prediction of 305-day lactation yields is crucial for dairy herd management and genetic evaluations.
• Existing methods may have limitations in handling incomplete or irregular test-day data.

Purpose of the Study:

• To develop and evaluate a multiple-trait procedure for predicting 305-day lactation yields (milk, fat, protein).
• To incorporate lactation curve information and covariances between yield traits into the prediction model.

Main Methods:

• Utilized standard lactation curves based on cow breed, region, lactation number, age, and calving season.
• Weighted test-day yields by their relative variances for parameter estimation.
• Compared the accuracy of the multiple-trait procedure with the test interval method.

Main Results:

• The multiple-trait procedure demonstrated comparable accuracy to the test interval method.
• The procedure effectively handles long test day intervals and single milk-only records.
• Predictions are possible from just one test day record per cow.

Conclusions:

• The developed multiple-trait procedure provides a robust method for predicting 305-day lactation yields.
• It offers valuable management insights, including peak yield, day of peak, and yield persistency.
• This method enhances the utility of milk recording programs for dairy producers.