4 April 1997

Trial aims to study environment/gene effects on proofs

Early results from a large Irish trial indicate that bull proofs made in countries where concentrate use is heavy are a reasonable guide to daughter performance in lower input systems. Robert Davies reports

HERD performance and feeding records from 608 farms on both sides of the border were made available to Andrew Cromie, a University of Dublin student completing his PhD project at The Agricultural Research Institute for Northern Ireland, Hillsborough, Co Down.

These were analysed to investigate whether the interaction between genetic make up and environment can cause the re-ranking of bulls for yield between high and low concentrate herds, or whether just the scale of the expression of the trait is changed.

Information on average concentrate input a cow a year for four years ending on Mar 31 last year was used to define herds on the basis on concentrate use. The bottom 25% of herds in terms of concentrate fed were placed in a low input group, and the top 25% in a high input group.

Over 23,000 individual cow records were made available, 12,463 in high input herds and 10653 in low input herds. About two-thirds of the animals in high input herds, and one-third of those in low input herds, were the progeny of imported Holstein sires.

Average concentrate input was 548kg a cow a year in low input herds, and 1576kg a year in high input herds. Herds getting higher rates of concentrate produced an average of 1400 extra litres a cow, and 100kg a cow of milk solids. Fat content was also higher, but protein was slightly lower. Calving interval was also 18 days longer. The annual performance of the different groups is given in table 1.

A total of 1049 sires were used by the 136 high input herds. By contrast, the 195 herds feeding an average of less than 1t of cake a cow used 522 bulls. Of the 230 sires common to both only 20 had proofs of sufficient reliability in both high and low input systems to give reliable estimates of regression and correlation between sire proofs.

Mr Cromie told a Hillsborough seminar on high genetic merit cows that the results suggested bull proofs from herds in Canada, the US and Holland that feed high levels of concentrate are reliable predictors of daughter performance in herds feeding lower levels in Ireland.

But interaction between genetic type and environment caused a scaling effect. The results showed that 60% of the perceived benefits from using sires proven in herds feeding high levels of concentrate were realised when the same sires were used in the low input systems.

The average performances of 13,095 heifers were used to investigate the genetic correlation for milk yield between high and low input herds. This, too, revealed little evidence of re-ranking. "Both the breeding value analysis and the genetic correlation analysis would suggest that, for the definitions of environments and sires present in this study, there is no evidence of re-ranking for milk production traits," he said.

"However it should be noted that the difference in environments of only about 1t a cow were not that marked. Comments on the reliability of bull proofs across countries based on these results should be treated with caution."

Further analysis would incorporate a much wider range of environments through access to UK data from National Milk Records and Genus. It was anticipated that this would enable a more conclusive esti-mate of genetic correlation between herds feeding different levels of concentrate to be established. &#42

Above: 60% of the perceived benefits of using sires proven in herds feeding high levels of concentrates were realised when the same sires were used in the low input system. Right: Andrew Cromie says there is little evidence of re-ranking of bulls.

Table 1. Annual performance a cow of high and low input herds for various traits

Average for each trait

TraitLow input herdsHigh input herds

Concentrate input (kg)5481,576

Milk yield (kg)4,7676,122

Fat yield (kg)178.3235.6

Protein yield (kg)156.0197.6

Fat %3.753.87

Protein %3.283.24

Calving interval (days)365383


&#8226 Proofs of bulls from high input countries reliable indicators of daughter performance in lower input herds.

&#8226 No evidence of re-ranking.