18 April 1997

Beef BLUP – its all or nothing

BEEF breeders must take on board full performance recording or not record at all if the accuracy of Signets beefbreeder service is to improve, according to a leading beef producer.

The service, which feeds the best linear unbiased prediction (BLUP) analysis used to calculate estimated breeding values, is in danger of being compromised because some breeders only partially record herds, says Richard Fuller, manager of Yorks-based JSR Farms.

"BLUP analyses do not benefit from enough complete data sets because too many breeders are either not recording at all or going for the cheap option of one weighing and, in many cases, no scanning."

Information going in

BLUP is only as good as the information going in, he says. "When that is not complete BLUP is having to make too many assumptions to fill in the gaps, which reduces accuracy of EBVs of some animals.

"Its so frustrating that partially recorded herds are challenging the integrity of those breeders who contribute to the system by participating and paying for full recording."

Full recording with three scanning visits costs about £4 a cow. "Thats less than the cost of two pints of beer and not even worth worrying about," he says. "But instead producers are just making do with the minimum needed for the service – one weighing visit – and that is no way near enough as all other performance records will then have to be predicted." The onus is on breeders to make full use of the service, he says.

Signet general manager John Southgate says he supports Mr Fullers pleas for more complete recording among beef breeders.

But he denies that the efforts of those who do record fully are being compromised by those recording only at the minimum level. "Every record helps towards the evaluation of animals within a breed. But those who do not record fully will have EBVs which are less robust with lower statistical accuracy."

MLC beef strategy manager Chris Brown is also supportive of full recording but doesnt wish to see it become obligatory. For some it may be impractical to record fully and the contribution of even minimally recorded herds will still improve the overall accuracy of the breed EBVs, he explains.

"Ideally, animals should be fully recorded, but Id rather see partial recording than nothing at all," says Dr Brown. His advice is dont worry too much about accuracy. "View it as an assessment of risk – and focus on EBVs and the level of those as a reflection of the genetic merit of an animal."