Hierarchical modeling of Baltic Sea phytoplankton reveals increasing cyanobacteria dominance

Abstract

Understanding how phytoplankton communities respond to environmental changes is critical for mitigating the impact of changing oceans. To identify the main drivers of changes in phytoplankton community composition and species richness, hierarchical modeling of species communities (HMSC) was used to model the summer phytoplankton communities from 29 sites, sampled between 1972 and 2017, in the Helsinki Archipelago, Finland. Models were fitted to both occurrence and abundance data. Spatial–temporal random effects accounted for most of the variance explained, followed by sea surface temperature (SST) and total phosphorus for both models. Statistically supported relationships between SST and occurrence emerged for most species. Two statistically supported trait–environment relationships were detected; however, species responses to environmental variables were strongly linked to taxonomic relatedness. A negative relationship between species richness and temperature was found. The results indicated that, independent of seasonal succession, the abundance of cyanobacteria was predicted to increase with increasing temperature, while the abundance of diatoms was predicted to decrease. This trend was most notable in the hurdle model, suggesting that even if species are not completely lost, there is still a shift in dominance from diatoms to cyanobacteria with increasing temperature. Additionally, higher total phosphorus levels also resulted in increased abundance of cyanobacteria. The primary environmental drivers of changes in the community composition of phytoplankton in the Helsinki Archipelago were SST and total phosphorus; however, the community is not temporally stable. The Baltic Sea is becoming warmer; therefore, summer phytoplankton communities are expected to be increasingly dominated by cyanobacteria.