A Global eDNA Comparison of Freshwater Bacterioplankton Assemblages Focusing on Large-River Floodplain Lakes of Brazil.

With its network of lotic and lentic habitats that shift during changes in seasonal connection, the tropical and subtropical large-river systems represent possibly the most dynamic of all aquatic environments. Pelagic water samples were collected from Brazilian floodplain lakes (total n = 58) in four flood-pulsed systems (Amazon [n = 21], Araguaia [n = 14], Paraná [n = 15], and Pantanal [n = 8]) in 2011-2012 and sequenced via 454 for bacterial environmental DNA using 16S amplicons; additional abiotic field and laboratory measurements were collected for the assayed lakes. We report here a global comparison of the bacterioplankton makeup of freshwater systems, focusing on a comparison of Brazilian lakes with similar freshwater systems across the globe. The results indicate a surprising similarity at higher taxonomic levels of the bacterioplankton in Brazilian freshwater with global sites. However, substantial novel diversity at the family level was also observed for the Brazilian freshwater systems. Brazilian freshwater bacterioplankton richness was relatively average globally. Ordination results indicate that Brazilian bacterioplankton composition is unique from other areas of the globe. Using Brazil-only ordinations, floodplain system differentiation most strongly correlated with dissolved oxygen, pH, and phosphate. Our data on Brazilian freshwater systems in combination with analysis of a collection of freshwater environmental samples from across the globe offers the first regional picture of bacterioplankton diversity in these important freshwater systems.

Description of freshwater bacterial assemblages from the upper paraná river floodpulse system, Brazil.

Bacteria were identified from a large, seasonally flooded river (Paraná River, Brazil) and two floodplain habitats that were part of the same river system yet very different in nature: clearwater Garças Lagoon and the highly humic waters of Patos Lagoon. Bacterioplankton were collected during mid-summer (Jan. 2002) from water samples (2 l) filtered first through a 1.2-microm filter then a 0.2-microm membrane filter representing the particle-attached and free-living sub-communities, respectively. DNA was extracted from filters and purified and a 16S rRNA clone library established for each habitat. Over 300 clones were sequenced and checked for similarity to existing 16S sequences in GenBank using the BLAST algorithm with default parameters. Further classification of clones was done using a species "backbone" attachment followed by parsimony analysis. The majority (85%) of sequences, referred to here as operational taxonomic units (OTUs), were most similar to uncultured bacterium 16S sequences. OTUs from each Proteobacteria sub-phylum (alpha, beta, gamma, delta, epsilon) were present in the Upper Paraná River system, as well as members of the Bacteroidetes. The microbial assemblage from Patos Lagoon was least like other samples in that it had no Firmicutes present and was dominated by Actinobacteria. Verrucomicrobia OTUs were only found in the free-living assemblage. This study documents the presence of globally distributed phyla in Upper Paraná River and taxa unique to habitat and particle attachment.

Culturability of stream bacteria assessed at the assemblage and population levels.

Lotic bacterial communities can be examined at multiple levels: from the assemblage level to populations of individual species. In stream environments, as in many other systems, the percentage of bacteria that are culturable is quite low. In this study, the culturability of the overall bacterial assemblage, as well as the culturability of three common species (Acinetobacter calcoaceticus, Burkholderia cepacia, and Pseudomonas putida), was determined in samples collected from four streams on three dates. Colony hybridization (colonies were grown on modified nutrient agar) and fluorescent in situ hybridization were used to calculate the percentage of cells of a given species that were culturable. Approximately half of the overall assemblage was estimated to be viable but nonculturable cells (VBNC). The culturability of two of the species was low (0.29% for A. calcoaceticus and 0.46% for P. putida), whereas the value for B. cepacia (2.48%) exceeded the overall assemblage level culturability (0.90%). Overall, both bacterial assemblages and populations were dominated by VBNC. These results show quantitatively that not all members of a species that has culturable representatives are culturable when retrieved from natural populations, likely because of interspecific phenotypic and genotypic variability. Thus, the large pool of nonculturable cells includes representatives of species that are, under some circumstances, culturable.