Selection for milk production and persistency using eigenvectors of the random regression coefficient matrix

Area of Science:

• Animal Breeding
• Quantitative Genetics
• Dairy Cattle Improvement

Background:

• Understanding genetic variation in lactation curves is essential for effective dairy cattle breeding.
• Additive genetic random regression models provide a framework for analyzing longitudinal traits like milk production.

Purpose of the Study:

• To investigate the relationship between eigenvectors of the additive genetic random regression coefficient matrix (K) and selection responses.
• To determine the optimal number of eigenvectors required in breeding goals for explaining variation in lactation traits.

Main Methods:

• Construction of eigenvector indexes using the K matrix estimated from test-day milk records of Japanese Holstein cattle.
• Analysis of genetic responses generated by different eigenvector indexes across lactation days.

Main Results:

• The first eigenvector scaled lactation curves without altering shape; subsequent eigenvectors influenced curve shape.
• The second and third eigenvectors significantly contributed to changes in lactation curve shape.
• An index using the first three eigenvectors captured over 99.9% of the genetic response for improving both milk yield and persistency.

Conclusions:

• Eigenvector indexes, particularly the first three, are highly effective for genetic selection in dairy cattle.
• The utility of higher-order eigenvectors (fourth and fifth) for improving milk yield or persistency is limited.
• Eigenvector indexes offer advantages over conventional selection, especially when persistency has a higher economic weight.