Characterization of bacterial communities in four freshwater lakes differing in nutrient load and food web structure.

The phylogenetic composition of bacterioplankton communities in the water column of four shallow eutrophic lakes was analyzed by partially sequencing cloned 16S rRNA genes and by PCR-DGGE analysis. The four lakes differed in nutrient load and food web structure: two were in a clearwater state and had dense stands of submerged macrophytes, while two others were in a turbid state characterized by the occurrence of phytoplankton blooms. One turbid and one clearwater lake had very high nutrient levels (total phosphorus > 100 microg/l), while the other lakes were less nutrient rich (total phosphorus < 100 microg/l). Cluster analysis, multidimensional scaling and ANOSIM (analysis of similarity) were used to investigate differences among the bacterial community composition in the four lakes. Our results show that each lake has its own distinct bacterioplankton community. The samples of lake Blankaart differed substantially from those of the other lakes; this pattern was consistent throughout the year of study. The bacterioplankton community composition in lake Blankaart seems to be less diverse and less stable than in the other three lakes. Clone library results reveal that Actinobacteria strongly dominated the bacterial community in lake Blankaart. The relative abundance of Betaproteobacteria was low, whereas this group was dominant in the other three lakes. Turbid lakes had a higher representation of Cyanobacteria, while clearwater lakes were characterized by more representatives of the Bacteroidetes. Correlating our DGGE data with environmental parameters, using the BIOENV procedure, suggests that differences are partly related to the equilibrium state of the lake.