A major challenge of contemporary marine science is disentangling consequences of climate change from other impacts, and studying non-target species and using historical resources to see long-term trends can meet this need. However, such data can be fragmented, and here, we demonstrate the potential of leveraging across sources for insight. We assembled a variety of historical sources such as scientific and personal observations, anecdotal information, and archival fisheries data to create an abundance time series on threespine stickleback Gasterosteus aculeatus in the White Sea starting in the late 19th century - the longest time series for this species. Stickleback peaked during the warm period of the 1920-1940s and declined during the colder period of the 1950-1990s and now is the most numerous vertebrate in the sea. Analyses of historical and recent time series based on our own data (2007-2019) showed that stickleback abundance decreases during colder winters. It is not associated with zooplankton biomass, positively correlated with herring Clupea sp. catches and negatively with navaga Eleginus navaga catches. Large population size and food web interactions suggest that change in stickleback abundance has the potential to affect the entire White Sea ecosystem.