摘要

It is important to assess ovulation detection performance in commercial dairy herds both to investigate low reproductive performance and to enable herd managers to monitor the effectiveness of their system for detecting ovulations. A method was developed to assess ovulation detection performance that uses limited numbers of strategically collected milk samples, assesses performance over the period when herd managers are making maximal effort to detect ovulations, and when assessing proportions of ovulations detected, accounts for false positive diagnoses of estrus and for cows that have not recommenced postpartum ovulatory cycles. Milk was sampled from cows not diagnosed in estrus early in the breeding program (about d 26 in year-round calving herds and d 22 in seasonal calving herds); milk samples were also collected from cows on the day of insemination. Cows with high milk progesterone concentrations were assumed to have had undetected ovulations and false positive diagnoses of estrus, respectively. The method was successfully implemented in 161 of 167 commercial dairy herds. Positive predictive values (PPV; the proportions of ovulation diagnoses where ovulation was, in fact, imminent) were generally high in both year-round and seasonal calving herds (median values were 0.96 and 0.97, respectively), but 25% of herds had PPV <0.95. Ovulation detection sensitivities (ODS) were low in most year-round calving herds, but many seasonal calving herds had high ODS values; median ODS were 0.73 and 0.94, respectively. However, in 25% of seasonal calving herds, ODS was <0.91. These findings indicate that this method for assessing ovulation detection performance can be successfully implemented in commercial dairy herds with appropriate professional support. The wide range of ODS and the absence of correlation between ODS and PPV suggest that it is possible for managers of many commercial herds in Australia to achieve increased reproductive efficiency through increases in ODS and PPV.

  • 出版日期2010-7