Is sperm DNA fragmentation a good marker for field AI bull fertility?

This paper aimed at investigating the potential use of sperm DNA fragmentation (SDF) to improve the routine screening of infertility of Holstein bulls. Cryopreserved sperm samples from 201 Holstein bulls provided by an AI center were used in the analyses of SDF at 0 (SDF_0) and 6 (SDF_6) h of incubation at 37°C. A refinement of the sperm chromatin dispersion test implemented in the Sperm-Halomax kit was employed to measure SDF. Records on routinely collected semen traits (volume, concentration, mass and individual motility evaluated in the fresh ejaculate, and individual motility in post-thawed semen straws) were provided by the AI center. Artificial insemination bull fertility was obtained from official field recording as successful or failed insemination. The results show that the average SDF was low (around 3.5%) at 0 and 6 h of incubation. A moderate effect of inbreeding depression was found. Estimated heritability for SDF traits were moderately high (0.41 and 0.29 for SDF_0 and SDF_6, respectively) and estimated repeatability of SDF measures in the same animal were high (0.73 and 0.70 for SDF_0 and SDF_6, respectively). An overall estimated service bull value (ESBV) obtained through statistical modeling that allowed for adjustment of systematic environmental effects not specific to a bull and of the female contribution to fertility, and the estimated genetic values (EGV) were obtained from field-recorded AI information. The ESBV and EGV were also obtained for all semen traits. Moderately large and negative Pearson correlation coefficients were observed between SDF traits and male fertility ranging from (-0.43 to -0.50; P <0.001). Results of stepwise regression analyses showed that SDF_6 had the largest partial r(2) (0.15 to 0.26) among all semen characteristics. Overall, the selected semen traits explained 25% and 31% of the observed variability in bull fertility measured as EGV and ESBV, respectively. When looking at the predictive ability of bull fertility categories, the results of discriminant and logistic regression analyses showed that low-fertility bulls (those in the 10th or lower percentile in the fertility distribution) can be accurately identified by using measures of SDF alone or in combination with sperm motility. Values of SDF around 7% to 10% could be used as indicators of low AI success.