Vertical distribution of ammonia-oxidizing crenarchaeota and methanogens in the epipelagic waters of Lake Kivu (Rwanda-Democratic Republic of the Congo)

Abstract

Four stratified basins in Lake Kivu (Rwanda-Democratic Republic of the Congo) were sampled in March 2007 to investigate the abundance, distribution, and potential biogeochemical role of planktonic archaea. We used fluorescence in situ hybridization with catalyzed-reported deposition microscopic counts (CARD-FISH), denaturing gradient gel electrophoresis (DGGE) fingerprinting, and quantitative PCR (qPCR) of signature genes for ammonia-oxidizing archaea (16S rRNA for marine Crenarchaeota group 1.1a [MCG1] and ammonia monooxygenase subunit A [amoA]). Abundance of archaea ranged from 1 to 4.5% of total DAPI (4',6-diamidino-2-phenylindole) counts with maximal concentrations at the oxic-anoxic transition zone (∼50-m depth). Phylogenetic analysis of the archaeal planktonic community revealed a higher level of richness of crenarchaeal 16S rRNA gene sequences (21 of the 28 operational taxonomic units [OTUs] identified [75%]) over euryarchaeotal ones (7 OTUs). Sequences affiliated with the kingdom Euryarchaeota were mainly recovered from the anoxic water compartment and mostly grouped into methanogenic lineages (Methanosarcinales and Methanocellales). In turn, crenarchaeal phylotypes were recovered throughout the sampled epipelagic waters (0- to 100-m depth), with clear phylogenetic segregation along the transition from oxic to anoxic water masses. Thus, whereas in the anoxic hypolimnion crenarchaeotal OTUs were mainly assigned to the miscellaneous crenarchaeotic group, the OTUs from the oxic-anoxic transition and above belonged to Crenarchaeota groups 1.1a and 1.1b, two lineages containing most of the ammonia-oxidizing representatives known so far. The concomitant vertical distribution of both nitrite and nitrate maxima and the copy numbers of both MCG1 16S rRNA and amoA genes suggest the potential implication of Crenarchaeota in nitrification processes occurring in the epilimnetic waters of the lake.

FIG. 1.

Geographical location of Lake Kivu and the four sampling stations: northern (NB), southern (SB), eastern (EB) and Bukavu Bay (BB) basins. The geographical GPS (global positioning system) coordinates for each sampling site are indicated. Basins are named as by Sarmento et al. (67, 68).

FIG. 2.

Depth profiles of temperature (solid line), conductivity (dashed line), and dissolved oxygen (open triangles) in the different basins on the day of sampling.

FIG. 3.

Relative abundances of Bacteria and Archaea obtained after CARD-FISH counts (left) and total cell numbers (DAPI staining, right) in northern basin (NB) and Bukavu Bay (BB). Error bars show the standard deviations from results for triplicate filters (see Materials and Methods for details).

FIG. 4.

Neighbor-joining phylogenetic tree generated by the ARB software package showing the affiliation of the OTUs retrieved from Lake Kivu (in bold). Bootstrap support values of >50% (1,000 replicates) are shown. The scale bar indicates 10% estimated sequence divergence. OTU identification numbers (OTU ID) correspond to the same numbers shown in Tables 1 and 2 in the supplemental material. Sequences after OTU ID correspond to the representative sequence for each OTU obtained with mothur (see Materials and Methods for details). Sequences are named according to the primer used (a or b for primer PAIRa or PAIRb, respectively), the basin (K for Kibuye [eastern basin], I for Ishungu [southern basin], G for Goma [northern basin], and B for Bukavu [Bukavu Bay]), and the band number for each basin (see Fig. S1 and S2 in the supplemental material).

FIG. 5.

Vertical distribution of ammonium, nitrate, and nitrite (left panels), gene abundance (MCG1 16S rRNA and crenarchaeal amoA) (middle panels), and DGGE fingerprints of amoA gene fragments (right panels) in the water column of eastern (EB, upper) and southern (SB, bottom) basins of Lake Kivu. Identical amoA fingerprints were obtained with samples collected from NB and BB stations (data not shown). Bands were named using the code for the corresponding basin (K for Kibuye [eastern basin, EB] and I for Ishungu [southern basin, SB]) and numbered sequentially from top to bottom of the gradient. *, unspecific products. Nitrite was undetectable in all depths from the eastern basin.

FIG. 6.

Neighbor-joining phylogenetic tree for amoA sequences constructed using Tamura-Nei-corrected distances with a bootstrap value of 1,000. Bootstrap values higher than 50% are shown. Wedge sizes are proportional to sequences condensed on them. Brackets highlight environmental clusters. The scale bar indicates a 5% sequence dissimilarity.