Southern elephant seal population breeding probability declines with increased sea-ice cover in a key foraging region

Regional differences in southern elephant seal (Mirounga leonina) population trends have long been hypothesised to reflect environmental variation affecting prey availability. By combining census counts, demographic data, and locally relevant winter sea-ice conditions, we provide new evidence quantifying the suspected link between seal numbers in the Macquarie Island population and regional sea-ice cover in the Southern Ocean. We developed an integrated population model using annual abundance and capture–mark–recapture data from 1980–2023, accounting for observation error and temporal random effects. Population breeding probability varied annually (50-65% on average). To assess how environmental conditions influence reproductive dynamics, we tested the effect of winter sea-ice cover within the known foraging range of adult females. We compared two covariate structures and a null model, incorporating one-year-lagged (B-COV1) and same-year (C-COV0) sea-ice conditions. Covariate models confirmed a negative relationship (βB-COV1 = -0.0103 [-0.025; 0.0037] to βC-COV0 = -0.0345 [-0.049; -0.021]), indicating that reduced population breeding probability was associated with higher-than-average sea-ice cover. Estimates of the total population size (N) mirrored the observed decrease in seal numbers from annual abundance counts. N was used to project historical scenarios representing persistently ‘best’, ‘neutral’, and ‘worst’ case sea-ice cover conditions over the full timeseries. The scenario-specific intrinsic growth rates ranged from a slow decrease under the best-case scenario (r = –0.495%) to a substantially stronger decline under the worst-case scenario (r = –3.126%), with the neutral scenario closely matching the observed population trend (r ≈ –2%). These comparisons emphasise how local-regional scale habitat conditions can alter long-term population trajectories at magnitudes relevant to species recovery.