Temporal Trends in Harbor Seal ( Phoca vitulina ) Counts Throughout Removal of Artificial Haul‐Out Sites Adjacent to Critical Salmon Habitat

ABSTRACT Pinnipeds have been blamed for the lack of recovery of fishes of human importance. Removing artificial haul‐out sites has been proposed to reduce pinniped presence and presumably predation. However, the effectiveness of this nonlethal management method has not been examined. We analyzed harbor seal ( Phoca vitulina ) numbers at two sites in Bellingham, WA, USA, where seals haul out on floating haul‐out structures adjacent to Whatcom Creek, used by Pacific salmon ( Oncorhynchus spp.). We used Generalized Additive Mixed Models ( GAMMs ) with a continuous‐time AR(1) correlation structure to predict seasonal and interannual trends in seal counts at the haul‐out sites and Whatcom Creek before (pre‐2015), during (2015–2022), and after (post‐2022) floating haul‐out removal. Model‐predicted haul‐outs at Log Pond peaked at 33.3 seals in 2013 (95% CI: 26.9–41.1), declined by 90% to 3.4 seals in 2017 (95% CI: 2.4–4.8), rose by 101% to 6.8 seals in 2019 (95% CI: 5.1–9.1), and declined to near zero seals in 2023 (95% CI: 0.0–0.0). At Creek, model‐predicted seal counts increased by 135% from 2.4 seals in 2011 (95% CI: 1.9–3.2) to 5.8 seals in 2021 (95% CI: 5.1–6.4). The 2011–2021 increase in model‐predicted seals at Whatcom Creek was followed by a 6% decrease to 5.4 seals in 2023 (95% CI: 4.6–6.4). These predominantly divergent trends between hauled‐out seals and seals swimming in Whatcom Creek suggest that removing artificial haul‐out structures may reduce the number of hauled‐out seals at targeted sites without necessarily reducing seal counts in adjacent foraging habitat.