A topic under intensive study in community ecology and biogeography is the degree to which microscopic, as well as macroscopic organisms, show spatially‐structured variation in community characteristics. In general, unicellular microscopic organisms are regarded as ubiquitously distributed and, therefore, without a clear biogeographic signal. This view was summarized 75 years ago by Baas‐Becking, who stated “everything is everywhere, but, the environment selects”. Within the context of metacommunity theory, this hypothesis is congruent with the species sorting model. By using a broad‐scale dataset on stream diatom communities and environmental predictor variables across most of Finland, our main aim was to test this hypothesis. Patterns of spatial autocorrelation were evaluated by Moran's I based correlograms, whereas partial regression analysis and partial redundancy analysis were used to quantify the relative importance of environmental and spatial factors on total species richness and on community composition, respectively. Significant patterns of spatial autocorrelation were found for all environmental variables, which also varied widely. Our main results were clear‐cut. In general, pure spatial effects clearly overcame those of environmental effects, with the former explaining much more variation in species richness and community composition. Most likely, missing environmental variables cannot explain the higher predictive power of spatial variables, because we measured key factors that have previously been found to be the most important variables (e.g. pH, conductivity, colour, phosphorus, nitrogen) shaping the structure of diatom communities. Therefore, our results provided only limited support for the Baas‐Becking hypothesis and the species sorting perspective of metacommunity theory.

The potential atmospheric impact of constructed wetlands (CWs) should be examined as there is a worldwide increase in the development of these systems. Fluxes of N(2)O, CH(4), and CO(2) have been measured from CWs in Estonia, Finland, Norway, and Poland during winter and summer in horizontal and vertical subsurface flow (HSSF and VSSF), free surface water (FSW), and overland and groundwater flow (OGF) wetlands. The fluxes of N(2)O-N, CH(4)-C, and CO(2)-C ranged from -2.1 to 1000, -32 to 38 000, and -840 to 93 000 mg m(-2) d(-1), respectively. Emissions of N(2)O and CH(4) were significantly higher during summer than during winter. The VSSF wetlands had the highest fluxes of N(2)O during both summer and winter. Methane emissions were highest from the FSW wetlands during wintertime. In the HSSF wetlands, the emissions of N(2)O and CH(4) were in general highest in the inlet section. The vegetated ponds in the FSW wetlands released more N(2)O than the nonvegetated ponds. The global warming potential (GWP), summarizing the mean N(2)O and CH(4) emissions, ranged from 5700 to 26000 and 830 to 5100 mg CO(2) equivalents m(-2) d(-1) for the four CW types in summer and winter, respectively. The wintertime GWP was 8.5 to 89.5% of the corresponding summertime GWP, which highlights the importance of the cold season in the annual greenhouse gas release from north temperate and boreal CWs. However, due to their generally small area North European CWs were suggested to represent only a minor source for atmospheric N(2)O and CH(4).

No abstract

No abstract