Sire Evaluation for Type in Holstein-Friesian Cattle

The appropriateness of the current system of calculating predicted difference type was compared to other methods including four methods of herdmate deviation and a daughter-dam method using final score type proofs for 331 Holstein-Friesian sires with a minimum of 150 officially classified daughters each. Cumulative groups of 10, 20,..., 50 daughters per sire (initial proof) and a sample containing the remaining classified daughters of each sire were used to calculate proofs. The later sample averaged 434 daughters per sire and intended to portray the genetic transmitting ability of the bull. Correlations between initial and later proofs ranged from .49 to .77 and were similar for the method presently used and the four herdmate methods. Correlations for the daughter-dam method were considerably lower. While the method currently used to calculate the predicted difference type by the Holstein-Friesian Association of America appeared to be as repeatable as any of the herdmate methods, correlation was sizable (.39) between the predicted difference type and the average of average scores of herds in which the bull was tested initially. This correlation approached zero for herdmate deviation which included an adjustment for the genetic type of dams (mates). Accuracy of ranking Holstein bulls for type could be improved considerably by changing to deviation from herdmates.