Predicting boar fertility using the sperm mobility assay: method optimization and early predictive modeling using logistic and machine learning approaches

Authors: Mills, Kayla; MINTON, AMANDA - Acufast; MAGEE, MAC - Resource Concepts; Long, Julie

Submitted to: Translational Animal Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 8/25/2025
Publication Date: N/A
Citation: N/A

Interpretive Summary: Accurately identifying subfertile boars before they are used for breeding is a major challenge in the swine industry. Traditional semen assessments like motility and morphology estimates lack the ability to accurately predict the rate at which a boar will produce offspring or litter size. The goal of this study was to adapt the sperm mobility assay for current boar stud workflows, a test that can predict fertility in poultry, and determine if the assay could predict fertility in boars. We modified the assay to work with regular breeding doses of semen used in commercial boar studs since the boar studs dilute the sperm cells to a standardized breeding dose concentration. When tested on semen from boars with known conception rates, the results showed that boars with high conception rates (>80%) had higher sperm mobility. We also found that the assay was able to pick up on differences in boar health weeks before the boars were experiencing any clinical symptoms that were reflected in lower mobility readings and translated to lower overall fertility rates for the stud experiencing a disease outbreak. This suggests the test might also help with early detection of hidden health issues that affect fertility. Overall, this low-cost and easy-to-use assay could give producers a better way to improve boar fertility and avoid the negative economic consequences associated with subfertile boars.

Technical Abstract: Subfertile boars remain a persistent challenge for the swine industry, and current methods for semen evaluation such as motility and morphology estimates do not reliably predict fertility outcomes which calls for the need for supplemental assessments capable of fertility prediction. The sperm mobility assay, originally developed for poultry fertility selection, uses a cell-separation solution known as Accudenz where absorbance values indicate the rate of sperm cell penetration into the solution and is correlated to fertility. Previous work validated the assay in boars based on concentration of viable sperm cells, but the assay’s predictive ability for fertility classification was never evaluated. Moreover, current stud workflows extend ejaculates that pass minimum semen quality thresholds for motility and morphology to a final breeding dose concentration. Therefore, the objectives of this study were to adapt the assay for semen extended to final breeding dose concentration and evaluate its predictive potential for boar fertility classification. In Experiment 1, four combinations of Accudenz concentration (3% or 6%) and volume (1.5ml or 3ml) were tested using extended semen from 16 commercial Duroc boars to identify conditions that produced mobility readings comparable to previous methods for boars. The 1.5 ml of 3% Accudenz assay, yielded the most comparable readings to previous studies and was used for determining predictive performance of the assay. Twenty-one commercial Duroc boars from two boar studs with known conception rate were enrolled in the study: High-conception rate (HCR; >80%, n=11) and (LCR; <75%, n=10). Overall, HCR boars consistently exhibited higher mobility readings and reached their maximum mobility values sooner than LCR boars indicating a greater ability to penetrate through Accudenz. Binomial regression using 5-minute absorbance readings effectively distinguished fertility groups across all boars (AUC = 0.740, P < 0.01). While the inclusion of stud as an interactive effect maintained model performance (AUC = 0.739), significance was reduced (P = 0.12), and within-stud models revealed variability in predictive strength (Stud A: AUC = 0.62; Stud B: AUC = 0.71). A random forest concluded that the inclusion of more animals into the existing model would strengthen its predictive ability, but the potential is still there. The mobility readings observed in Stud B was the main driver of the model’s predictive ability. The uniqueness of Stud B’s absorbance patterns and great magnitude between HCR and LCR mobility readings could be attributed as an extender effect or a severe health challenge that may have contributed to reduced sperm viability and mobility, suggesting potential use of the assay as a sentinel for subclinical reproductive disease. Overall, the sperm mobility assay shows promise as a low-input, commercially feasible tool to complement existing semen evaluation methods and improve boar fertility prediction.