Some Factors Associated With Fertility of Thoroughbred Stallions

The results of 3 years (2005–2007) of observations and mating (5,646 estrous cycles of 3,788 mares bred to 1 of 15 stallions) at one Thoroughbred breeding farm in central Kentucky were analyzed by a multiple logistic regression model using Bayesian statistics to evaluate the relationship between data entries (factors) and pregnancy outcomes. Factors found to be significantly (P < .05) associated with pregnancy outcome included stallion (one stallion had lower OR for pregnancy higher odds ratio OR] for pregnancy, and one had, than other stallions), date of mating (OR for pregnancy declined slightly in May – July), mare age (OR for pregnancy were higher for mares <13 years old, and lower for mares >18 years old), mare beginning status (foaling mares had a higher OR for pregnancy), mating on foal heat (lowered OR for pregnancy), mating of the day for the stallion (OR for pregnancy was 4.16 times lower for fifth compared with first mating of day), reinforcement breeding (increased OR for pregnancy), dismount semen neutrophil score (lowered OR for pregnancy when neutrophils were present in dismount semen samples), and tranquilization before breeding (lowered OR for pregnancy in foaling and barren mares). The influence of dismount sample sperm motility scores on OR for pregnancy was weak, so motility scores were not included in the final logistic regression model. The majority of variation in pregnancy outcome was because of mare factors, with only approximately one-third of the variation in fertility explained by stallion.