Abundances of breeding seabirds as indicators of the environmental state in the Baltic Sea – challenges and alternative approaches

• Direct use of seabird abundances as indicators can provide misleading information. • Seabird abundance may be affected by pressures outside the focal area, e.g. inland. • The indicator’s focal area should match the spatial scale of population dynamics. • Transient dynamics and species interactions can be filtered out from indicators. • Modelling techniques help to aggregate multi-species signals and filter out noise. Policy-relevant environmental indicators are needed to support effective management of marine environments. Here, we present common challenges in selecting a set of seabird indicator species, whose abundances are used as indicators, and propose potential solutions. We use examples from the Baltic Sea and emphasize the requirements of the EU’s Marine Strategy Framework Directive (MSFD). Processes that confound indicators can be classified as 1) contradictory indirect anthropogenic effects, e.g. poor state of Baltic cod stocks increasing auk abundances, 2) drivers affecting seabird abundance but not the state of the sea, e.g. decreased availability of dumps for gulls, and 3) transient dynamics and species interactions, e.g. the steep population growth of Great Cormorant and White-tailed Sea Eagle, and the resulting predation impact of the latter on Common Eider. To assess the spatial scale of seabird population dynamics, we compared population trends of the Baltic Sea with the North Sea, and trends of the Finnish coast with inland lakes. On average, trends in the Baltic Sea and Finnish coast were more stable or positive compared to the North Sea and inland, respectively. Finally, we propose two phases of indicator formation: i) a filtering phase – in which unwanted variation is filtered out from the species’ population trajectories – and ii) an indicator-generating phase – which aggregates relevant multi-species patterns into one or several indicators. We suggest statistical methodological approaches that cover both phases, separately or simultaneously. At best, seabird abundance indicators are useful summary statistics for monitoring joint effects of a wide range of anthropogenic pressures.