A multiple-trait herd cluster model for international dairy sire evaluation

Area of Science:

• Animal Genetics
• Dairy Science
• Agricultural Economics

Background:

• Traditional international dairy sire evaluations use a two-step process, transforming national breeding values across countries.
• This method is limited by defining environments solely by country borders, ignoring similarities between neighboring countries and differences within large countries.

Purpose of the Study:

• To propose and evaluate an international genetic evaluation model using herd clusters.
• To overcome limitations of country-based evaluations by grouping herds with similar environmental and management characteristics.

Main Methods:

• Utilized 4.6 million lactation records from 46,000 herds across multiple countries and US regions.
• Clustered herds into five groups based on 13 descriptive variables including herd size, management practices, and environmental factors.
• Calculated genetic correlations between herd clusters, ranging from 0.81 to 0.97.

Main Results:

• Successfully formed five distinct herd clusters, each encompassing herds from multiple countries or regions.
• Demonstrated high genetic correlations between clusters, indicating the model captures meaningful genetic similarities.
• Reduced the number of traits from 13 to 5, simplifying the evaluation process.

Conclusions:

• The herd cluster model provides a more intuitive and accurate prediction of animal genetic merit within unique environments, irrespective of national boundaries.
• This parsimonious and computationally feasible model is suitable for large-scale international genetic evaluations.
• The findings support a paradigm shift from country-specific to environment-specific genetic evaluations in the dairy industry.