Genomics and NZ Animal Evaluation Did they get it right?

We’ve had a full season of evaluation that incorporates genomic data from one company only into NZ’s animal evaluation system - so has it lived up to the hype?
 
First of all, we want make a couple of things clear - CRV AmBreed’s position on the use of genomics in Animal Evaluation in the New Zealand Dairy industry has not changed since NZAEL introduced it back in May 09.  Genomic Selection is a powerful technology that has the potential to significantly increase rates of genetic gain for dairy farmers, improving productivity and overall profitability of dairy farm operations in NZ.
 
Having been at the forefront of the development of this technology globally, CRV has invested heavily in genomics in NZ and around the world, and believes that in the near future it will show significant benefits for farmers. 
 
The results from the Interbull validation tests show that NZAEL’s decision to include Genomics in the NZ Animal Evaluation system in 2009 may have been too early for its stage of development- as shown by the variability in validation test results.
 
In the opinion of CRV, the advice given to farmers by NZAEL over-simplifies the issues around the over-estimation of young sires, and may lead farmers to believe that the problem NZAEL have identified with genomic data on young sires applies equally to all young sires.   The recommendation of NZAEL to make an across the board allowance of 15-25BW is a “normal” range expected of Parental Average overestimation, but analyses of the overestimation of the highest ranked bulls with genomic data is showing a range of 60-70BW.
 
The movements of these top-ranked bulls are of real importance to the industry.   CRV’s study into the Re-Proof performance of top bulls shows that in NZ up to 90% of all replacement heifers from one year group of bulls can be sired by just the top 6 bulls of the two leading genetics companies.  The ramifications of ‘getting it wrong’ with genetic merit estimation are obviously significant.
 
CRV AmBreed has not supplied its genomic data on young sires for inclusion in Animal Evaluation because we do not believe these issues have been scientifically resolved, and we do not believe it is wise or appropriate to test the system on NZ dairy farmers.
 
We have been providing genomically selected young sires through our InSire product range since 2008 – and have at all times urged our clients to exercise caution by balancing the risk profile of the genetics being used.
 
The results from the Interbull validation are encouraging in relation to the state of the technology.  The current problem lies firstly in that some traits perform well while others perform poorly and secondly in the early incorporation of genomic data from one company into Animal Evaluation and the subsequent over-hyped “all-in” commercial promotions that were run here in NZ.
 
Contrast this approach with that taken by CRV in Holland, where they will incorporate genomic data into their national animal evaluation system in the coming months.
 
In Holland, CRV’s approach to incorporating genomic information has differed in three significant areas:

• The key to accurate genomic data is the size of the reference population.  CRV waited until it had over 16,000 bulls from one breed in its reference population compared to approximately 3,000 bulls across three breeds being used in NZ.
• CRV will include no dam information from the dam side.  This removes the troublesome parental average bias from the genomic estimations.
• They waited until internationally accepted validation processes were made available to give greater confidence in the results.

 
We congratulate NZAEL on the openness with which they have conducted this review, and we call for the three issues facing Animal Evaluation to be clearly addressed separately.   Those three issues are:
 

• Parental Overestimation
• Bias in Genomic Data
• The Reproof effect

 
 Genomics will play a valuable role in the future of animal breeding in New Zealand and within CRV, and we will continue our research and test programs in this area.  CRV AmBreed continues to recommend the PA/PT derived index as the most scientifically reliable and proven resource for bull selection and to treat genomic BV’s with the appropriate degree of caution until such time as scientific rigour stabilises the BW and reliability values.