Rod Polkinghorne

LIVESTOCK industry consultant, Dr Alex Ball has added his support to a call for a ‘Virtual Red Meat Data Hub’, with the potential to link animals recorded on separate databases such as NLIS, Livestock Data Link, MSA, Sheep Genetics and beef cattle breed society databases.

“Nearly all of the carcase data used for in genetic analysis in the future will come from commercial animals, so it is imperative that we can link carcase data from commercial animals back to the genetic analysis,” Dr Ball told last week’s Livestock Breeding and Genetics forum in Brisbane.

There was no need for commercial breeders to record the sire and dam of each animal, as this could be done with genomics technology, he said.

“I have done an analysis of the bulls used compared to the bulls that could have been used in one breed, and I estimate that  the losses in the beef value chain by not being able to directly link consumer requirements to bull breeders is about $5.5 million dollars per year. Who is paying for that loss?” Dr Ball asked.

Supporting his call, stalwart beef industry consultant Dr Rod Polkinghorne said it was astounding that of the MSA carcase data collected on two million animals per year, carcase records on only about 500 animals were used in the Breedplan analyses for the Australia seedstock industry.

Organised by Meat and Livestock Australia, last week’s Brisbane forum was surprisingly strongly attended, attracting 350 producers, researchers and industry service providers, creating extensive discussion around the need for a national livestock data platform.

Currently there was no linkage between databases such as the Meat Standards Australia database, Livestock Data Link (LDL) and databases such as the National Beef Recording Service (NBRS) and Sheep Genetics (SG) which are used for genetic analysis of beef and sheep seed stock.

The hope is that carcase measurements for commercial animals captured on the MSA and LDL databases could be used in an expanded Breedplan and Sheep Genetics Genetic analyses, but this depends on the use of unique identification numbers across the databases and the ability to establish the actual sires of these commercial animals.

The National Livestock Identification System provides unique identification numbers for all slaughter cattle, but because it is voluntary for sheep, only very limited numbers of sheep are slaughtered. However the NLIS identification is generally not recorded on seedstock animals. Identification of the sires of these commercial animals requires a genomic parentage test on individual slaughter animals so that the sires can be established.

Irish model examined

During the Brisbane forum, the Irish Cattle Breeding Federation (ICBF) database was promoted by several speakers as a model for the Australian beef industry, but there were some views that this may not be a practical model for Australia.

The Irish approach to genetic gain is covered in detail in this earlier article on Beef Central.

The ICBF model has been built with large financial contributions from the European Union which enables the payment of 90 Euros (about A$ 120) per animal which is recorded on the database by the owner of the beef or dairy animal.

As well as recording the birth date, sire and dam and at least one weight on the animal, the owner is also required to submit a hair sample for genomic testing. Carcase measurements taken at slaughter by the abattoir are submitted to the database using a national identification number. However it was pointed out that the ICBF system does not measure eating quality of beef.

A subset of the recorded data is then used to calculate EBVs and simple Indexes for beef and dairy seedstock animals. There is an incentive for commercial breeders to use semen from top ranked ‘five star’ sires, based on a breeding index incorporating calving ease, fertility, growth and carcase traits.

But because of small herd sizes and the lack of recording of different management practices for each animal on each farm, most of the data collected is of little value for genetic evaluation (calculation of EBVs and indexes).

There is little evidence that genetic improvement achieved by the Irish beef and dairy industries is achieved by genetic improvement by Irish breeders. Instead, the vast majority of genetic gain achieved in Ireland is imported.

In beef it has been achieved via imported Limousin and Charolais semen and bulls mainly from France and in the case of dairy from other countries, according to a prominent Australian geneticist.

The Irish achievement is much more an intelligent use of EU funds to put their cattle industry on an information footing – by having all the different purpose databases talking to each other, than anything substantial about genetic gain.

“They have certainly driven adoption or use of genetic information and developed new (very simple) indexes that are helping commercial producers buy better genetics for their commercial operation,” the geneticist said.

The Australian beef industry generates a far higher proportion of genetic gain from within our populations – so the selection purpose of BreedPlan is generating at least as much wealth as the herd bull choice purpose – and much more over the medium-term, he said.

“One possible long-term benefit of the fascination with Ireland might be that industry realises and acts on the fact that real genetic gain is worth investing in – much more than has been to date and doing so in a coordinated fashion,” he said.

New Zealand approach

Andrew Cooke, a principal of New Zealand based information technology company, Rezare Systems, outlined to the Brisbane audience the elements of a Data Platform that might be suitable for Australian needs.

“It must be able access data from multiple sources, and have controlled access for data sharing. It can have a wide variety of applications, and should provide powerful analytics,” Mr Cooke said.

Rezare Systems has built a data platform for Beef & Lamb New Zealand (roughly the equivalent of MLA) to replace a functional, but ageing system. The platform stores genotypes, phenotypes and pedigrees for multiple species.

“The core components are databases, an integration layer, or ‘API’, which allows access for a range of data providers and has processing services,” he said.

One unanswered question from the forum was who will build the national livestock data platform for Australia. The most obvious body to do this would be MLA, which owns or controls most of the key data bases that need to be linked under the proposed model.