Environmental drivers and zooplankton functional composition
The environmental variables that have the potential to shape zooplankton communities (salinity, temperature, turbidity), as well as the proxies of eutrophication (chlorophyll a , dissolved oxygen) were not measured consistently during the monitoring cruises. The model data are therefore considered a realistic, and a more accurate (without measurement error) estimate of salinity, oxygen and temperature. Oxygen, temperature and salinity data originate from the 3D coupled sea-ice ocean model for Baltic Sea (BSIOM; Lehmann et al. 2014), which has a high spatial resolution (2.5 × 2.5 grid cells) and a 3 m depth resolution for the period of 1979-2014. Oxygen concentration was extracted from the model for the bottom layer, and salinity and temperature were calculated as average values of water column, matching the zooplankton observations (latitude, longitude, month). Winter air temperatures were obtained from three weather stations in the north-eastern part of Gulf of Riga ( Pärnu, Kihnu and Sõrve). The Baltic Sea Index (BSI; provided by A. Lehmann, Helmholtz Centre of Ocean Research), defined as the difference of normalized sea level pressure (SLP) anomalies between Oslo in Norway and Szczecin in Poland (Lehmann et al. 2002), was used as a proxy for large-scale atmospheric variability. To investigate the potential connections between environmental variables and zooplankton community composition, we used a hypothesis-driven approach.