Thursday, June 1, 2017

BIF 2017: Implications of multi-breed evaluations and across-breed EPDs for commercial cattlemen

Bruce Golden
Theta Solutions

IGS Analysis:

  • 12 breed associations
  • 12 million pedigree records
  • 10 million observation records
  • 15 different EPDs
The IGS primary goals are:
Improve accuracy
EPDs that can be compared between breeds
Weekly run
Advanced methods

In order for the data to be comparable across breeds, there has to be connectedness between pedigrees. There have to be sires that are in Simmental pedigrees that are also in Red Angus and Gelbvieh pedigrees. With all of these breeds using Angus sires to produce hybrids, there is a good possibility that there will be good linkages between breeds.

Models seem to be robust to methods used or not used to adjust for heterosis. Have to adjust for heterosis effects, as they are not additive, to effectively predict EPDs (additive effects).

Larry Kuehen
USDA-MARC

The center uses 2,200 AI matings each year that result in about 1,000 AI-sired pregnancies. 

Adjustment factors will be coming in July of these year. Factors are also published with performance breed averages.

USDA-MARC wants to get to the point where all breeds in the IGS evaluation have the same common adjustment factor to get to the Angus scale.

The group is also working towards releasing adjustment factors multiple times a year. 
They are also working towards releasing across-breed EPD factors in early spring, before bull buying season.

Joe Epperly
North American Limousin Foundation

We have ~35 million calves produced each year in the U.S. beef industry. This is kind of like a salad of different centuries. A bit of this and a bit of that. However, in the seedstock industry, we look at each individual ingredient. 

While breeds have different averages, the range in performance between breeds overlaps. 

Epperly encourages use of planned matings tools made available by breed associations.

Matt Spangler
University of Nebraska-Lincoln

Why is there a desire for EPD comparable across breeds?
This helps facilitate informed decisions by commerical cattlemen.

A multi-breed does not inherently mean that EPD are directly comparable across multiple breeds.
Direct comparisons require:
  • Pedigree ties
  • Common base adjustments
  • Common trait definitions
  • Common underlying models
  • Correct accounting for breed and heterotic effects
Additive adjustment factors from USDA-MARC are needed if:
  • Breed and heterotic effects cannot be reasonably estimated from field data due to confounding of contemporary group and breed.
  • A common base is not adopted.
  • Etc.
The goal is to have all IGS partner breeds with directly comparable breeds. This assumes that each IGS partner breeds will use, adopt, and publish IGS EPDs and not breed-specific EPDs.

Q&A
Don't compare EPDs of bulls across breeds, unless we first use the adjustment factors. You will know breeds are directly comparable when the USDA-MARC adjustment factors for the breeds are identical. 

Separate software from models. BOLT is simply software. We have to have the correct analytical models for issues to be corrected.

At the end of the day, the question is-are we ranking sires correctly? If this answer is yes with a simple model, then a complex model is not necessary.


Post a Comment