The value of the percentage of motile sperm in predicting a significant portion of the fertility variation of frozen-thawed buck semen

Prediction of the future fertility of a given ejaculate with a simple laboratory test is still considered a real issue in domestic mammal breeding. This study showed that a subjective assessment of the percentage of motile spermatozoa, measured 120 min after thawing (mob120), can predict a significant part (∼50%) of the variation of the future fertility of buck ejaculates. The predictive model was calculated using a calibration data set composed of 40 ejaculates from four Alpine and six Saanen bucks. A fertility trial using split ejaculates was conducted in order to estimate ejaculate fertility. Taken into account were the herd within breed factor and the year, month, and inseminator factors. On average, one ejaculate was used to inseminate two females per herd in 10 different herds. This calibration set allowed us to choose the mob120 variable among a set of laboratory tests: mitochondrial activity, acrosomal status, membrane integrity, osmotic resistance test assessed by flow cytometry, velocity and motion characteristics assessed by computer-assisted sperm analysis, visually assessed percentage of motile, and motility score measured 5 and 120 min after thawing. For the calibration step, the best model used the logarithm of mob120 and gave a correlation coefficient of 0.71 between the field fertility and the predicted fertility and a standard error of 0.17. We tested this model on 3 different validation data sets adding up to 95 ejaculates that were all different from those of the calibration data set. The correlation coefficients between field fertility and predicted fertility were always significant and the bias corrected standard error ranged from 0.15 to 0.18 on these validation data sets. A Monte Carlo simulation showed that about 20% of the fertility variation remained to be explained.