Seabird responses to altered marine vessel activity during the COVID-19 anthropause: insights from citizen science

Human activities exert multifaceted pressures on marine ecosystems, yet the degree to which seabirds respond to changes in vessel activity remains poorly quantified. The COVID-19 "anthropause" provided an unprecedented natural experiment to assess these responses across spatial and temporal scales. We combined Automatic Identification System data describing marine vessel activity with citizen-science observations from eBird to examine how 35 coastal seabird species in British Columbia, Canada, responded to pandemic-related changes in vessel traffic. Vessel activity was modelled using Generalized Additive Models to identify sites where small (personal), mid-sized (fishing and tug), and large (cargo, tanker and passenger) vessel traffic increased or decreased relative to pre-pandemic years (2018-2019). Seabird abundance was then modelled using negative binomial Generalized Additive Mixed Models to test for species-level changes between pre- and post-COVID periods, using the same sites identified in the previous step as showing significant increases or decreases in vessel activity. This paired-model approach (one for increase sites and one for decrease sites) allowed us to isolate seabird responses to directional changes in vessel activity, increasing confidence that observed patterns reflected vessel effects rather than unrelated environmental variation. Results revealed strong spatial heterogeneity in vessel reductions, with large vessels showing widespread declines, while small and mid-sized vessels displayed mixed patterns. Seabird responses were closely linked to these patterns: abundance generally increased where small-vessel traffic declined and decreased where it increased, indicating a negative association with small boats, while the opposite was true for mid-sized vessels, with abundance tending to rise where traffic increased and fall where it declined, reflecting a positive association. Large vessels produced few, mostly positive, significant associations, suggesting weaker or more distant effects. Collectively, these results demonstrate that vessel type, predictability, and proximity influence seabird responses to maritime disturbance. Our findings underscore the importance of considering vessel class and species-specific behaviour in marine spatial planning and mitigation of human impacts on seabird communities.