Search Results

You are looking at 1 - 1 of 1 items for

  • Author: Robert John Aitken x
  • Refine by Access: Open Access content only x
Clear All Modify Search
Open access

Robert John Aitken, Sarah Lambourne, and Ashlee Jade Medica

In brief

A capacity to predict the likelihood of pregnancy following natural matings would be of considerable benefit to the Thoroughbred horse breeding industry. In this article, we describe a strategy for achieving this outcome through the analysis of dismount samples, that achieved an overall accuracy of 94.6%.


The purpose of this study was to determine whether the analysis of dismount semen samples from Thoroughbred stallions could be used to predict whether a given mating would result in a pregnancy. The analysis was based on 143 matings of 141 mares by a cohort of 7 Thoroughbred stallions over a 4-week period at an Australian Stud. The criteria of semen quality utilized in this analysis involved a preliminary assessment of the raw dismount sample in terms of semen volume, sperm number, and sperm movement characteristics using an iSperm® Equine portable device. Following this initial assessment, a subpopulation of progressively motile spermatozoa was isolated by virtue of the cells ability to migrate across a 5 µm polycarbonate filter in a Samson™ isolation chamber over a 15-minute period. These isolated cells were again evaluated for their number and quality of movement using the iSperm® system and, in addition, assessed for their ability to reduce WST-1, a membrane impermeant tetrazolium salt. These data were then combined with additional information describing the ages of the animals used in this study, their historical breeding records, and mating frequency during the breeding season. The total data set was then used to predict the occurrence of pregnancy, as confirmed by ultrasound at ~14 days post mating. The criteria used to predict fertility in the ensuing multivariate discriminant analysis were optimized for each individual stallion. Using this strategy, we were able to successfully predict the outcome of a cover with an overall accuracy of 94.6%.