Macrosystem community change in lake phytoplankton and its implications for diversity and function

Aim: We use lake phytoplankton community data to quantify the spatio-temporal and scale-dependent impacts of eutrophication, land-use, and climate change on species niches and community assembly processes while accounting for species traits and phylogenetic constraints.
Location: Finland
Time period: 1977-2017
Major taxa: Phytoplankton
Methods: We use Hierarchical Modelling of Species Communities (HMSC) to model meta-community trajectories at 853 lakes over four decades of environmental change, including a hierarchical spatial structure to account for scale-dependent processes. Using a ‘region of common profile’ approach, we evaluate compositional changes of species communities and trait profiles and investigate their temporal development. 
Results: We demonstrate the emergence of novel and widespread community composition clusters in previously more compositional homogeneous communities, with cluster-specific community trait profiles, indicating functional differences. A strong phyloge..., The National Finnish Phytoplankton Monitoring Database is maintained by the Finnish Environment Institute (SYKE; open data portal http://www.syke.fi/en-US/Open_information) and comprises nationwide phytoplankton community data of lake surface water samples. 
We used data collected in the late summer months with samples taken from early July to late August, reflecting the peak productivity season of lake phytoplankton communities. To ensure consistent sampling methodology, we included only data between the years 1977 and 2017. All phytoplankton samples were preserved with acid Lugol’s solution and analysed using the standard Utermöhl technique.
As detailed in the manuscript this data supports, we included only species that contributed to the cumulative upper 95% of the total biomass during the four decades of monitoring. This resulted in a total of 165 species, for which we here provide sample-specific presence-absence data.
Our study area includes 853 boreal lakes in Finland with 1057 d..., All analysis was performed in R version 3.6.3 (R Development Core Team, 2020) and is open-source. The provided script includes all necessary steps of opening and structuring the data, as well as fitting and post-processing of the models to reproduce the results of the manuscript.Â