Ordered categorical sire evaluation for dystocia in Holsteins

Area of Science:

• Animal Science
• Biostatistics
• Genetics

Background:

• Ordered categorical data, like calving ease scores, present analytical challenges.
• Traditional methods may not fully capture the complexities of such data.
• Accurate analysis is crucial for genetic improvement and animal welfare.

Purpose of the Study:

• To investigate new statistical procedures for analyzing ordered categorical data.
• To apply the standardized threshold model to calving ease data.
• To estimate fixed and random effects on an underlying continuous scale.

Main Methods:

• Utilized the standardized threshold model for ordered categorical data analysis.
• Modeled calving ease using sex of calf, parity of dam, and unknown boundary points as fixed effects.
• Included herd-year-seasons and sires as random effects in a dataset of 5029 dystocia scores.

Main Results:

• Male calves were predicted to be twice as likely to experience difficult births compared to females.
• Difficult births decreased by a factor of 5 between the first and second parity.
• Heritability estimates were higher on the underlying scale (.147) than the observed scale (.061).

Conclusions:

• The standardized threshold model provides a robust framework for analyzing ordered categorical data in animal breeding.
• Statistical insights into calving ease can inform breeding strategies and improve animal welfare.
• Herd-year-seasons and sire effects significantly influence calving ease, with heritability being higher on the underlying continuous scale.