Semen fertility--an evaluation system for artificial insemination sires, technicians, herds, and systematic fixed effects.

Semen fertility was evaluated on 4478 artificial insemination bulls using 2,007,520 breeding receipts for inseminations to 397,979 Holstein cows between January 1, 1978 and November 1, 1984 using mixed linear model methodology. Bulls differed on average by +/- 3.68% conception. Differences between herd-years ranged from 27.4 to -32.3% conception with an average difference of +/- 8.33%. Within herds, cow parity-service number accounted for the largest variation, ranging from 24.0% to -18.5% conception. June was the best month for conception (1.9) and January the poorest (-2.3), indicating effect of temperature and confinement in winter on estrous detection in the northeastern United States. Cows with intervals between repeat services of 21 d had conception rates 12% higher than intervals between repeat services of 11 d. Effect of month-year of semen collection varied by +/- 1.33% conception. Differences between technicians within years varied by +/- 3.37%. Results of the analysis indicate improved conception on dairy farms can be achieved by improved estrous detection, which is a major component of herd-year and service interval. Other components of the model contributed little to the variance in conception or are uncontrollable by farm management.