Sire by region interaction for production traits in beef cattle

Data from 38,261 field records obtained from the American Simmental Association were used to evaluate the importance of sire x region interactions in beef cattle for birth weight, weaning weight, yearling weight, gestation length and calving ease.Data were analyzed by least squares procedures by a model containing sire, region and sire x region interaction effects.Trait ratios were used in all analyses.Only progeny born and weaned in the same herd were included.The United States was divided into nine regions with one region deleted due to small numbers.Montana was further divided into three sub-regions to study the interaction within one state for birth weight and weaning weight.Birth weight and weaning weight were analyzed using separate analyses of records from all eight regions, four regions containing 85% of the records, two regions (Texas and Oklahoma vs. Montana) and the three regions of Montana.Records from five regions were analyzed for yearling weight and from three regions for gestation length and calving ease.The number of records for these latter three traits was small.The progeny of 10-13 sires were used in each analysis.The region effect was not significant for any trait.The sire effect was significant for all analyses except weaning weight within Montana, yearling weight and calving ease.The sire x region interaction effect was significant for all analyses of weaning weight (P<.01) except Montana and approached significance (P=.07) for birth weight within Montana.Genetic correlations were calculated from the sire and sire x region interaction variance components for each analysis to assess the average change in sire performance among regions.Correlations were not defined for the within Montana weaning weight or calving ease.Correlations of unity were obtained for all analyses of birth weight except Montana and for gestation length indicating no change in the ranking of sires.For birth weight within Montana and yearling weight, correlations of 0.21 and 0.25 were obtained, respectively, suggesting some changes in the ranking of sires.Correlations of 0.47 to 0.54 suggested that changes in the ranking of sires occurred for weaning weight.This was confirmed from the least squares means for the sire x region subclasses.The sine x region variance component was less than 27° of the total variance for any given analysis of weaning weight.Although this was a small percent, it could be important biologically.The least squares means for the sire x region subclasses indicated that in many cases a given sire ranked above average in one region but below average in other regions.This indicated that errors could be made in applying the results of progeny tests in one region to progeny performance and selection in a different region.The need for evaluating sires in all regions and on a within region basis was suggested.