Swimming against time: Temporal dynamics of sperm motility in Arctic charr (Salvelinus alpinus)

Arctic charr (Salvelinus alpinus) exhibits highly variable reproductive performance compared to other salmonids. Sperm motility is a critical determinant of fertilisation success, yet temporal changes in motility patterns during the breeding season remain poorly understood. We assessed sperm quality during commercial roe production by sampling 49 males over three consecutive weeks, tracking changes in ten distinct motility subpopulations throughout the activation period using Computer-assisted sperm analysis (CASA). We compared the effects of two activation media: tank water and the commercial activator used at the facility (AquaBoost®) on motility subpopulation dynamics, examined whether phenotypic scores influence sperm swimming patterns, and investigated whether different ovarian fluids trigger varying sperm responses from the same males. Sperm concentration averaged 3.19 ± 1.91 billion sperm/mL with no significant weekly variation. Sperm concentration emerged as the dominant predictor of activation success (odds ratio = 3.10 per 10⁹ sperm/mL). AquaBoost® activator significantly extended motility duration by 6.5 s compared to tank water and was associated with an increased proportion of rapid-swimming sperm, while tank water was associated with higher proportions of medium-velocity categories. Cochran-Mantel-Haenszel analysis revealed significant temporal shifts in motility subpopulations, with Medium Super Progressive sperm declining consistently across weeks (θ = 2.66 for AquaBoost®, θ = 2.44 for water, where θ > 1 denotes higher odds in the earlier week). The decline was confirmed at the male level and was concentrated immediately after activation (odds ratio 7.98 at 0-4 s) and disappeared by 30 s. Progressive motility at 10-14 s declined across the sampling period for both activators. Phenotypic scores showed no detectable effects. Ovarian fluid effects were female-specific and temporally restricted. Male × ovarian fluid interactions were suggestive in five motility classes during the first 25 s post-activation but did not survive correction for multiple testing (q = 0.054-0.076) and are presented as exploratory. Our findings demonstrate that temporal subpopulation analysis might reveal patterns masked by average motility parameters, providing new insights relevant for both natural spawning and artificial fertilisation protocols.