Calculation of BREEDPLAN EBVs for North American Sires

November 23, 2016 10:52 am

Over the past six months, queries have been raised by several Angus Australia members regarding the calculation of EBVs for overseas sires in the Angus BREEDPLAN genetic evaluation, particularly at the time of, or shortly after importation.

This article discusses how EBVs are calculated for imported North American sires, reasons why Angus BREEDPLAN EBVs and North American EPDs can differ for an individual sire, the results of a recent study that evaluated the reliability by which starting EBVs are assigned for imported sires, and several initiatives that have been instigated to further improve the calculation of EBVs for North American sires.

 

Calculation of EBVs for North American Sires

The calculation of EBVs for sires imported from North America is quite challenging, particularly until such time as the sires have progeny with performance information recorded in Australia and New Zealand. At the time of importation, sires often have considerable pedigree, performance and genomic information recorded in North American genetic evaluations, yet very limited information available in Australia and New Zealand.

To address these difficulties, the Angus BREEDPLAN analysis utilises a number of strategies when calculating EBVs for recently imported sires, including the formation of specific genetic groups for animals of North American origin, and incorporation of EPD information from a number of overseas genetic evaluations.

 

Assigning a Starting EBV Value

When animals are first included in the Angus BREEDPLAN analysis, the animals are assigned a starting EBV value based on the EBVs of their sire and dam (known as a mid-parent EBV). If the animal’s sire and dam does not have EBVs, as is often the case for imported sires, the new animals are assigned a starting value that reflects the origin of their genetics and the expectation of these genetics, as determined by similar animals in the analysis (called a genetic group).

In Angus BREEDPLAN, different genetic groups are formed based on the animal’s year of birth (<= 1980, 1981-1990, 1991-2000, >=2001), country of origin (Australia, New Zealand, North America, United Kingdom/Other) and breed (British, European, Bos indicus, Other).

In situations where the new animals have been evaluated in another genetic evaluation (e.g. a North American genetic evaluation), Angus BREEDPLAN enhances the starting EBV value of these animals by importing the EPDs/EBVs from the other evaluation.

Angus BREEDPLAN currently utilises EPD information from the American Angus Association, the Canadian Angus Association, the Red Angus Association of America and American Brangus Association, with EPDs imported for birth weight, weaning weight, yearling weight, mature cow weight, milk, scrotal circumference, carcase weight, rib eye area, rib fat depth and marbling.

The imported EPD information is converted into local progeny equivalents for the traits of interest, with the local progeny equivalents being used to adjust the starting EBV value of the introduced animals around the average of their genetic group. That is, rather than all animals in a genetic group being assigned the same starting EBV, the animals’ overseas EPDs are used to spread the starting EBVs around the average of their genetic group.

These strategies reflect the methodology by which starting EBVs are assigned to imported North American sires. As performance information is subsequently recorded in Australia and New Zealand for the imported sire’s progeny, the sire’s EBVs will change to reflect this information, so that over time, the EBVs will be based on the performance of the sire’s progeny in Australia and New Zealand, rather than the sire’s North American EPDs and the “starting” value it was assigned.

 

Differences between EPDs and EBVs

While it would generally be expected that the North American EPDs and Angus BREEDPLAN EBVs for a sire will provide a similar indication of the sire’s breeding value for each trait, it is important to recognise that there are several reasons why the EBVs of a sire in Angus BREEDPLAN may differ to the EPDs of the sire in North American genetic evaluations.

Differences in EBVs and EPDs are not necessarily an indication of an analytical software error in one of the respective analyses, but rather are more likely a function of different information being analysed through a different analytical model. More specifically:

  • The traits for which EBVs and EPDs are calculated are different. For example, the carcase EBVs in Angus BREEDPLAN are expressed as differences in carcase traits in a standard weight carcase, while the carcase EPDs in the North American genetic evaluations reflect differences in carcase traits at a standard age.
  • The analytical models used in Angus BREEDPLAN and North American genetic evaluations are different. For example, Angus BREEDPLAN predominantly uses one main multi-trait model where all traits are considered simultaneously, while the North American genetic evaluations use a series of smaller multi-trait models where only smaller groups of traits are analysed simultaneously.
  • The genetic parameters utilised in Angus BREEDPLAN and in North American genetic evaluations are specific and relevant to the respective countries (e.g. trait heritabilities, trait correlations, adjustment factors).
  • The genomic predictions utilised by the genomic companies in North America and Australia/New Zealand are different and relevant to the respective country, resulting in the potential for the same animal to have vastly different genomic breeding values included in the North American genetic evaluation and in Angus BREEDPLAN.
  • The analytical model used to incorporate genomic information in the calculation of breeding values is different in Angus BREEDPLAN and the North American genetic evaluations, and the relative emphasis given to genomic information versus pedigree and performance information is different in Angus BREEDPLAN versus the North American genetic evaluations.
  • There is different information utilised in the calculation of EBVs in Angus BREEDPLAN and in the calculation of EPDs in overseas genetic evaluations. The Angus BREEDPLAN EBVs will reflect the information that is recorded in Australia/New Zealand, whereas the North American EPDs will reflect the information is recorded in North America.
  • Sires may perform differently in Australia/New Zealand compared with how they perform in North America, with the EBVs and EPDs reflecting how their progeny have performed in the respective countries.

 

Change in EBVs for Recently Imported North American Sires

Following several queries from members over recent months as to the calculation of EBVs for imported North American sires, Angus Australia and the Agricultural Business Research Institute (ABRI) have conducted a number of test analyses to evaluate the change in EBVs that have been observed for North American sires whose first progeny were born in Australia/New Zealand in 2010, 2011 & 2012, and who now have greater than 50 performance recorded progeny in Angus BREEDPLAN. This represented approximately 70 sires.

These analyses have revealed that while the EBVs for some individual sires have changed considerably since the time of importation, on average, there is no significant change in EBVs between when a sire was first imported and after such time as there is greater than 50 performance recorded progeny in Angus BREEDPLAN.

This analysis would suggest the current strategies being utilised within Angus BREEDPLAN are on average, providing a reliable indication of the breeding values for these sires at the time of importation.

Included below is a table below outlining the average EBVs when the sires were first imported in 2010, 2011 & 2012, and the current EBVs of the sires based on the performance of their progeny in Australia and New Zealand. Also illustrated in figures 1, 2 and 3 is the change in EMA, Rib Fat and IMF EBVs for each individual sire during this period.

 

CoEBVSFIS_Table

 

figure 1

 

figure 2

 

figure 3

 

Improvements to the Calculation of EBVs for North American Sires

While the investigation into the EBVs for imported sires has confirmed the starting EBV values are on average, being appropriately assigned to sires imported from North America, a number of initiatives have been instigated that aim to further improving the genetic evaluation of North American sires in Angus BREEDPLAN. These include:

  • A review will be conducted by the Animal Genetics and Breeding Unit (AGBU) into the overseas genetic evaluations from which EPDs are imported, the traits for which EPDs are imported, and the parameters that are used when importing EPDs in Angus BREEDPLAN.
  • A review will be conducted by ABRI of the genetic groups that are formed within the Angus BREEDPLAN analysis, and in particular, whether sufficient animals are now present to enable formation of additional genetic groups for North American animals.
  • Research will be conducted by AGBU into whether alternative and superior methodologies may be possible to calculate the EBVs for recently imported sires. This is particularly important with relation to the utilisation of genomic information on North American sires.
  • Angus Australia has commenced discussions with both the American Angus Association and Canadian Angus Association regarding opportunities for greater collaboration in genetic evaluation. This ranges from more frequent and affordable access to North American EPDs, to opportunities for across country genetic evaluation.
  • A number of North American sires will be included in future cohorts of the Angus Sire Benchmarking Program (ASBP) to generate linkage between the ASBP and American Angus progeny test programs. For example, G A R Prophet and A A R TEN X 7008 S A will be included in cohort 7 of the ASBP.

 

Summary

The calculation of EBVs for sires imported from North America is challenging with sires often having little or no performance information recorded in Australia and New Zealand at the time of importation. While on average, the starting EBVs assigned to North American sires provide a reliable indication of their breeding value, it should be expected that there will be some change in the EBVs of these sires until such time as there is sufficient performance information recorded for their progeny in Australia and New Zealand. A recent study conducted by Angus Australia and ABRI has confirmed the change in EBVs that may be observed for some individual sires is not a reflection of any bias or systematic error in the Angus BREEDPLAN analytical software, but rather is a result of the limited availability of information on which EBVs can be calculated at the time of, or shortly after importation.

 

Advertisements