OUP user menu

Bacterioplankton community diversity in a maritime Antarctic lake, determined by culture-dependent and culture-independent techniques

D.A. Pearce, C.J. van der Gast, B. Lawley, J.C. Ellis-Evans
DOI: http://dx.doi.org/10.1016/S0168-6496(03)00110-7 59-70 First published online: 1 July 2003


The biodiversity of the pelagic bacterioplankton community of a maritime Antarctic freshwater lake was examined by cultivation-dependent and cultivation-independent techniques to determine predominant bacterioplankton populations present. The culture-dependent techniques used were direct culture and observation, polymerase chain reaction amplification of 16S rRNA gene fragments, restriction fragment length polymorphism (RFLP) analysis followed by selective sequencing and fatty acid methyl ester analysis. The culture-independent techniques used were 16S ribosomal DNA gene cloning, RFLP analysis and sequencing, in situ hybridisation with group-specific, fluorescently labelled oligonucleotide probes and cloning and sequencing of dominant denaturing gradient gel electrophoresis products. Significant differences occurred between the results obtained with each method. However, sufficient overlap existed between the different methods to identify potentially significant groups. At least six different bacterial divisions including 24 genera were identified using culture-dependent techniques, and eight different bacterial divisions, including 23 genera, were identified using culture-independent techniques. Only five genera, Corynebacterium, Cytophaga, Flavobacterium, Janthinobacterium and Pseudomonas, could be identified using both sets of techniques, which represented four different bacterial divisions. Significantly for Antarctic freshwater lakes, pigment production is found within members of each of these genera. This work illustrates the importance of a comprehensive polyphasic approach in the analysis of lake bacterioplankton, and supports the ecological relevance of results obtained in earlier entirely culture-based studies.

  • Antarctic freshwater lake
  • Bacterioplankton
  • Community structure
  • 16S rRNA
  • Fatty acid methyl ester
  • Cloning

1 Introduction

The freshwater lakes on Signy Island, in the South Orkney Islands (60°43′S, 45°38′W), lie at the northern extremity of the maritime Antarctic. These lakes provide a latitudinal gradient of environmental conditions (temperature, light and nutrient levels) and complexity in their microbiology (diversity of organisms). There are 16 recognised lakes on Signy Island in a range of different catchment types, all of which are ice-covered to a depth of 1 m for 8–9 months each year. Each of the lakes has low species diversity and relatively short food chains, but tend to be more complex than systems further south. Lake temperatures range from 0 to 6°C and the nutrient content of individual lakes varies considerably, largely as a result of animal activity in accessible lakes near the sea.

Previous studies on the microbiology of Antarctic freshwater lake systems have used classical techniques, including direct identification using microscopy, or the growth of microorganisms under specific controlled conditions ([16], and references in reviews [7,8]). Indeed, only 10 years ago, our knowledge of bacteria that naturally occur in freshwater ecosystems was restricted to organisms that could be grown in culture [9]. The limitation of these methods, however, is that they rely upon either visual differences or upon selection of appropriate growth conditions. Culture-based approaches, while extremely useful for understanding the physiological potential of isolated organisms, do not necessarily provide comprehensive information on the composition of microbial communities [10], and ecological inferences based on the metabolic properties of cultivated bacteria are, by necessity, unrepresentative of the natural populations from which they were obtained. It is widely believed that culture methods alone are inadequate for studying microbial community composition [11], as colony morphology cannot always differentiate between genetically distinct isolates [12,13]. In addition, many microorganisms in the aquatic environment are not culturable [14], and with the recent interest in the ecological role of viable but non-culturable organisms [15], doubts have arisen concerning the ecological relevance of many cultures obtained from environmental samples. The potential for cultivable organisms to enter a viable but non-cultivable state can also dramatically distort the relative abundance of an organism observed in a culture collection [16].

The inadequacy of traditional cultural methods alone in describing a microbial community has been highlighted in a number of studies of widely different environments, where culture-independent molecular methods have been compared with culture-based methods [1728]. It is now generally accepted that our knowledge of bacterial diversity has been severely limited by the need to obtain pure cultures of microorganisms prior to their characterisation, when often <5% of bacteria in nature can be cultured with currently available methods [1214,29,30]. The lack of information is, in part, due to the very low biomass in these environments, ranging from <104 to 106 cells g−1[12,13]. However, since the inception of DNA-based identification methods, major taxa of previously unidentified organisms have been found in a wide variety of habitats. Dunbar et al. [16] cite numerous studies which investigate the phylogenetic overlap between organisms obtained by cultivation, and organisms identified by direct amplification and cloning of 16S ribosomal DNA (rDNA). They also cite studies in which no overlap was observed between culture collections and 16S rDNA clone libraries [31]. In another case, 41% of the phylotypes identified in a culture collection could also be identified in 16S rDNA clone libraries. They conclude that these studies have consistently demonstrated that the two methods generally sample different fractions of bacterial communities.

The ‘culturability’ problem is a general and well recognised dilemma for microbial ecologists regardless of the environment in question. Bharathi et al. [32] retrieved higher numbers of anaerobic bacteria than aerobic bacteria from Antarctic lake water samples, suggesting that many of the cells were in a viable but non-culturable state. Indeed, recent studies on 16S rDNA amplified directly from nucleic acids that have been extracted from natural environments have demonstrated that the vast majority of bacterial species are uncultured [33]. In addition to this, culture-based techniques can select for organisms without regard to their numerical or functional significance in situ [34]. Culture-based approaches, therefore, while extremely useful for understanding the physiological potential of isolated microorganisms, do not necessarily provide comprehensive information on the composition of microbial communities. To date, much of the ecological data from Antarctic lake systems are still based solely on cultured isolates.

Sombre Lake water was collected in February 2000 and the bacterioplankton subjected to a range of both culture-dependent and culture-independent identification techniques. The data obtained were then compared to existing data in the literature in an attempt to determine their ecological relevance.

2 Materials and methods

2.1 Sample site

Sombre Lake is situated in Paternoster Valley, on the north west corner of Signy Island in the South Orkney Islands (Fig. 1), and was selected as it is potentially the most studied lake in the maritime Antarctic. The lake is 250 m long and 150 m wide, with an area of 24 300 m2 and a maximum depth of 11.5 m. It is 5 m above mean sea level, 50 m from the sea and is ice-covered for up to 9 months each year. The drainage basin is composed of approximately 73% ice, 1% snow, 12% quartz-mica-schist, 1% amphibolite, 1% marble, 2% moss and 9% lichen cover, 2% seal-derived organic matter, and up to 2% bird-derived organic matter [35,36].


The location of Signy Island in the Southern Ocean.

2.2 Preliminary analysis of the bacterioplankton community

In a culture-based study in the 1980s, water samples were obtained at various depths in the water column, with sterilised water sampling bottles and returned in cool storage to the laboratory. A 10× dilution series was prepared for each sample and aliquots then spread-plated on casein-peptone-starch (CPS) medium [37]. Plates were incubated at both 3°C (for 22 days) and 10°C (for 10 days) and triplicate plates counted. Where required, representative colonies were then sub-cultured through liquid and solid media before storing pure strains at 3°C in 0.2%‘sloppy’ CPS media in screw cap bijou bottles for return to the UK, and long term preservation on beads in glycerol at −80°C.

2.3 Bacterioplankton direct culture

200 μl of lake water collected over the austral summer (1999/2000) was taken from a dilution series of lake water and placed onto solid CPS, TBSA and R2A plates in three replicates. Plates were incubated at 4°C for 2 months.

2.4 Fatty acid methyl ester (FAME) analysis

The phenotypic diversity and identity of individual strains isolated by direct culture were determined by FAME analysis, as described by Thompson et al. [38] and van der Gast et al. [39]. The samples were injected into a Hewlett-Packard model 5890 series II gas chromatograph, fatty acid peaks named by the Microbial Identification System (MIS) software (Microbial ID, Newark, DE, USA) and isolates identified using the MIS ‘Aerobe Library’.

2.5 Community DNA extraction

Sub-samples of water (1 l) were filtered through 0.2-μm cellulose nitrate filters (Whatman, Maidstone, UK) using a 250-ml filter unit (Sartorius, Goettingen, Germany). The filters were placed in a 30-ml sample tube (Sterilin, Feltham, Middlesex, UK) with 18 ml of the original unfiltered lake water sample, which was then mixed for 3 min. The resulting suspension was divided into 12 1.5-ml Eppendorf tubes and centrifuged for up to 60 min at 14 000 rpm (Force 14 Microcentrifuge, Denver Instruments, Denver, CO, USA). The pellets from each of four Eppendorf tubes were combined in triplicate and re-suspended in 1 ml lake water, which was then centrifuged for a further 60 min. The resultant pellet was re-suspended in 10 μl of lake water and subjected to five, 5-min freeze/thaw cycles alternating between −80°C and 40°C. Five microlitre of the supernatant was removed from immediately above the cell debris for use in the polymerase chain reaction (PCR).

2.6 PCR amplification of 16S rRNA gene fragments

Enzymatic amplification of the 16S rDNA was performed on DNA extracted directly from cultures, with primers 8F [16] (5′-AGAGTTTGATCCTGGCTCAG-3′) and 1500R [64] (5′-AGAAAGGAGGTGATCCAGCC-3′) using the method described in Pearce [40] and Pearce and Butler [41]. All DNA extraction procedures and manipulations were carried out in a laminar flow hood to minimise aerial contamination and all plasticware and equipment were exposed to 254 nm UV radiation for 15 min in a UV crosslinker (UVTech, Cambridge, UK) prior to use. All PCR products (10-μl volumes) were analysed by electrophoresis in 2% (w/v) agarose gels before further analysis was performed.

2.7 rDNA clone libraries

Prokaryote 16S rRNA genes were amplified by PCR using the near full length 8F and 1500R primer pair. For these amplifications, each PCR mixture (50 μl) contained 10 ng of extracted DNA as a template, 10 pmol of each primer, 20 nmol of each deoxynucleoside triphosphate, 1 U of SuperTaq polymerase (HT Biotech, Cambridge, UK) and the SuperTaq buffer supplied with the enzyme. Amplification reactions were performed with a Genius thermocycler (Techne, Minneapolis, USA) using the following conditions: an initial denaturation step consisting of 94°C for 5 min, 30 cycles consisting of 94°C for 45 s, 55°C for 45 s, 72°C for 70 s, and a final elongation step consisting of 72°C for 5 min. DNA free controls were also used to ensure that contaminants were not being amplified. The PCR products were cleaned using GFX PCR clean up columns (Pharmacia, NJ, USA). Cleaned products were ligated into the pGEMT-Easy vector (Promega, WI, USA) and ligation mixtures were transformed into competent JM109 cells as recommended by the manufacturer. Transformants were screened using black/white selection on Luria agar containing S-gal/IPTG and 50 μg ml−1 ampicillin (Sigma, St. Louis, MO, USA). Between 80 and 100 putative positive colonies were transferred to individual tubes containing 50 μl of sterile water. The cell suspensions were subjected to two freeze/thaw cycles and 1-μl aliquots were used as templates in a PCR reaction containing the M13F/M13R primer set (M13F 5′-CGCCAGGGTTTTCCCAGTCACGAC-3′ and M13R 5′-GAGCGGATAACAATTTCACACAGG-3′) and using the conditions above, except that the annealing temperature was raised to 58.5°C. Amplified products were separated electrophoretically and those containing inserts of the correct size were cleaned through GFX PCR clean up columns.

2.8 Restriction fragment length polymorphism (RFLP) analysis and sequencing

Aliquots of cloned rDNA were subjected to separate enzymatic digestions for 3 h at 37°C with the restriction enzymes BsuRI and Csp6I (MBI Fermentas, Vilnius, Lithuania). PCR product (5.0 μl) was placed into a 0.2-ml Eppendorf tube with 10 μl 10× enzyme buffer, 3.6 μl water and 0.4 μl of the respective restriction enzyme. The digestion products were electrophoresed in a 3% high-resolution agarose gel (Appligene, Illkirch Graffenstaden, France) for 2 h at 100 V. A 1-kb plus ladder (Helena Biosciences, TN, USA) was included in the gel to allow for subsequent normalisation. Gel images were analysed using Gelcompar II software (Applied Maths, Kortrijk, Belgium) and clones producing identical patterns with both enzymes were grouped into discrete operational taxonomic units (OTUs). At least one clone that was representative of each OTU was sequenced with both the M13F and M13R primers using the Big Dye terminator kit v.2 (Applied Biosystems, CA, USA). Sequence reactions were carried out in the Department of Biochemistry, University of Cambridge, and were run on an ABI Prism 3700 DNA analyser (Applied Biosystems, CA, USA). Clone sequences were compared with the GenBank nucleotide data library using gapped-blast searches (at http://www.ncbi.nlm.nih.gov/blast/blast.cgi) [42] to determine their closest phylogenetic neighbours.

2.9 Fluorescence in situ hybridisation (FISH)

The total population density of bacteria in water samples collected over a depth profile in Sombre Lake was determined by epifluorescence microscopy, using the 4′,6-diamidino-2-phenylindole (DAPI) staining technique [43]. Immediately after collection, 10 ml of water from each depth was fixed with buffered glutaraldehyde, to a final concentration of 2%, for enumeration of bacteria. The next day the fixed samples were filtered onto 0.2-μm black polycarbonate membranes (Poretics, Livermore, CA, USA) with an 8-μm backing filter to improve cell distribution, under a low vacuum (<50 mm Hg). Filters were stained with the sterile filtered fluorochrome DAPI for 5 min at a concentration of 2–5 μg ml−1, mounted onto slides and stored frozen at −20°C until enumerated by epifluorescence microscopy. Filters were mounted in Citifluor (Citifluor Products, Canterbury, Kent, UK) to minimise fluorescence fading during DAPI counts and viewed at 1250× magnification, under oil immersion, with a Leitz Labalux epifluorescence microscope fitted with a 50-W mercury lamp.

Ten millilitres of each lake water sample was then filtered through a black polycarbonate 0.2-μm screen membrane filter (Poretics, Livermore, CA, USA). Cells were fixed with 2 ml of 4% para-formaldehyde in phosphate-buffered saline for 30 min. A gentle vacuum was then applied and cells were rinsed, initially in 5 ml phosphate-buffered saline, then in 5 ml distilled water. Filters were removed from the filtration apparatus, air-dried, placed on a glass microscope slide and stored at −20°C. All preservation and hybridisation conditions were selected to minimise impact on the integrity and characteristics of cells.

Sixteen millilitres of hybridisation buffer containing one of five different in situ hybridisation probes was added to each filter and incubated at 46°C for 90 min in a closed hybridisation tank. The following oligonucleotide probes were used (numbering in parentheses according to Brosius et al. [44]): (i) EUB 338, complementary to a region of the 16S rRNA (338–355) specific for the domain Bacteria (5′-GCTGCCTCCCGTAGGAGT-3′) [45]; (ii) ALF 1b, complementary to a region of the 16S rRNA (19–35) conserved in the α-subclass of Proteobacteria and some other Proteobacteria (5′-CGTTCGYTCTGAGCCAG-3′) [46]; (iii) BET 42a, complementary to a region of the 23S rRNA (1027–1043) specific for the β-subclass of Proteobacteria (5′-GCCTTCCCACTTCGTTT-3′) [46]; (iv) GAM 42a, complementary to a region of the 23S rRNA (1027–1043) conserved in the γ-subclass of Proteobacteria (5′-GCCTTCCCACATCGTTT-3′) [46]; (v) CF 319a, complementary to a region of the 16S rRNA (319–336) conserved in the CytophagaFlavobacterium group (5′-TGGTCCGTGTCTCAGTAC-3′) [47]. Each probe was covalently linked at the 5′-end to a single fluorescent dye molecule – the very photostable indocarbocyanine dye CY3 [48]. The atmosphere in the hybridisation tank was saturated by placing excess hybridisation buffer on filter papers at the base of the tank. Each slide was then rinsed with 20 ml pre-warmed washing buffer at 48°C over a period of 15 min. Six hybridisation buffers were prepared with 0.9 M NaCl, 20 mM Tris–HCl (pH 7.4), 0.01% SDS and 50 ng ml−1 of one of the probes. Formamide was added to the following probes: ALF 1b, BET 42a, GAM 42a and CF 319a in concentrations of 20, 35, 35 and 15% respectively. Six washing buffers were also prepared with 20 mM Tris–HCl (pH 7.4), 5 mM EDTA, 0.01% SDS and NaCl in the following concentrations: 0.9 M EUB, 0.225 M ALF 1b, 80 mM BET 42a, 80 mM GAM 42a and 80 mM CF 319a. The filters were placed on slides, air-dried, mounted with the glycerol containing mountant Vectashield (Vector Laboratories Ltd, Peterborough, UK) to minimise bleaching and viewed at 1250× magnification, under oil immersion, with a Leitz Labalux epifluorescence microscope equipped with a 50-W mercury lamp and a CY3 filter set (filter set 41007A, Chroma, USA).

2.10 Denaturing gradient gel electrophoresis (DGGE)

A further set of amplifications was conducted with community DNA using the primers and amplification conditions described by Pearce [40] and Pearce and Butler [41]. Twenty-five microlitre of the amplification products were separated by DGGE (based on [49]). The linear denaturing gradient of urea and formamide, which ranged from 0% to 50% and which was 20 cm in length and 1 mm thick, was established (where 0% denaturant consisted of 6.5% 37.5:1 acrylamide:bisacrylamide mix in 1× TAE and 100% denaturant consisted of 6.5% 37.5:1 acrylamide:bisacrylamide mix, 40% formamide and 7 M urea in 1× TAE). The gel was run for 80 min at 60°C and 10 V cm−1, before staining for 45 min in 0.5 μg ml−1 ethidium bromide solution. This was visualised on a UV transilluminator (UVP, Cambridge, UK). Photographs were taken with a Gelcam (Polaroid, Cambridge, USA) using Polaroid 665 professional positive/negative instant pack film and exposure times of 20, 35 and 60 s, with a view to obtaining one over-saturated image, and one under-saturated image of the same gel, thereby increasing the information retrievable from one gel.

2.11 DGGE product clone libraries

The remaining 25 μl of PCR product from the DGGE samples was then cleaned using GFX PCR clean up columns (Pharmacia, NJ, USA). Cleaned products and excised, reamplified and cleaned gel bands were ligated into the pGEMT-Easy vector (Promega, WI, USA) and ligation mixtures were transformed into competent JM109 cells as described for the rDNA clone libraries, following RFLP analysis.

3 Results

3.1 Preliminary analysis of the bacterioplankton community

Using classical identification methods, eight groups of bacteria were identified from Sombre Lake, which is in agreement with numbers found in other lakes on Signy Island [50]. γ-Proteobacteria were predominantly isolated, mainly fluorescent Pseudomonads followed by non-fluorescent Pseudomonads and Vibrio, Aeromonas and the Moraxella/Acinetobacter groups. Isolated Bacteroidetes groups comprised of Flavobacterium and Cytophaga. The β-Proteobacteria Alcaligenes/Achromobacter and Chromobacterium (Janthinobacterium) and the Actinobacteria Micrococcus and coryneform bacteria, were isolated much less frequently [7]. The β-Proteobacterium Nitrosomonas sp. and the α-Proteobacterium Nitrobacter sp. were also isolated but neither showed significant activity at in situ temperatures [50] or was present in significant numbers. Chromogenic colony counts for Sombre Lake are given in Table 1.

View this table:

Sombre Lake bacteria

All values are %Chromogenic colonies – totalColor
Top water (1985)
Mid water (1987)
Bottom water (1985)
  • % chromogenic colonies on 3°C CPS plates. White and cream colonies were excluded. Orange colonies –Flavobacterium, Cytophaga, coryneform bacteria. Yellow colonies –Flavobacterium, Cytophaga, Vibrio, Micrococcus, coryneforms, Pseudomonads. Pink colonies – Pseudomonads. Violet colonies –chromobacterium (Janthinobacterium). Nd – none detected.

3.2 Direct 16S rDNA sequencing from cultured isolates

Two hundred aerobic, heterotrophic bacteria were isolated and categorised into 46 distinct groups based on colony morphology and RFLP analysis of 16S rRNA genes. Using direct amplification of 16S rDNA genes and sequencing, 72% of the bacteria were identified as β-Proteobacteria, 22% as Firmicutes, <1% as Actinobacteria, <1% as α-Proteobacteria and <1% as γ-Proteobacteria. The specific genera found were Aeromonas, Arthrobacter, Aquaspirillum, Bacillus, Desulfosporosinus, Duganella, Herbaspirillum, Janthinobacterium, Paenibacillus, Pseudomonas, Sporosarcina, Sphingomonas, Variovorax and a previously uncultured β-Proteobacterium.

3.3 FAME analysis

Using FAME analysis 49% of bacteria were identified as Firmicutes, 17% as Actinobacteria, 22% as γ-Proteobacteria, 9% as β-Proteobacteria and <1% as α-Proteobacteria. The specific genera found were the Actinobacteria genera Arcanobacterium, Arthrobacter, Cellulomonas, Kocuria, Microbacterium, Micrococcus and Rhodococcus, the firmicute genera Bacillus, Brevibacillus, Exiguobacterium, Paenibacillus, Staphylococcus, the α-Proteobacterium genus Brevundimonas, the β-Proteobacterium genus Janthinobacterium and the γ-Proteobacterium genera Flavimonas, Pseudomonas and Chryseomonas.

3.4 Clone sequences of 16S rRNA genes

Clone sequences of 16S rRNA genes contained 26% Actinobacteria, 25% Bacteroidetes, 6%α-Proteobacteria, 19%β-Proteobacteria, 3% Spirochaetales and 6% of δ-Proteobacteria, γ-Proteobacteria, and Verrucomicrobia. These included the Actinobacteria genera Clavibacter, Actinoplanes and Modestobacter, the Bacteroidetes genera Cytophaga, Flavobacterium and Flexibacter, the β-Proteobacterium genera Achromobacter, Bordetella, Thiobacillus, Acidovorax, Candidatus and Janthinobacterium, the α-Proteobacterium genera Sphingomonas and Rhodobacter, the γ-Proteobacterium genera Pseudomonas, the δ-Proteobacterium genera Desulfotalea and Pelobacter, the Verrucomicrobium genera Verrumicrobium and the Spirochaetales genus Spirochaeta.

3.5 FISH

The total DAPI-stained cell counts and the number of cells detected with probe EUB 338 are given in Fig. 2. These values represent hybridisation levels of between 48.6% at 1 m and 67.0% at 9 m. With a set of five probes for the major divisions within the domain bacteria, it was possible to affiliate between 54.1 and 69.7% of the EUB 338 hybridised cells with known bacterial groups. The β-Proteobacteria were clearly the most abundant group at all depths sampled and the γ-Proteobacteria were the least abundant group at all depths studied. The CytophagaFlavobacterium and α-Proteobacteria groups were intermediate between the β- and γ-Protobacteria numbers, with the CytophagaFlavobacterium group more common than the α-Proteobacteria (Fig. 3).


Number of cells detected using FISH with universal probe EUB 338 and DAPI staining over depth profiles in Sombre Lake. Values are means of 10 replicates±1 S.D.


Number of cells detected using FISH with group-specific probes over depth profiles in Sombre Lake. Values are means of 10 replicates±1 S.D.

3.6 DGGE

DGGE analysis of Sombre Lake water at all vertical depth intervals studied consistently produced the same eight distinct DGGE bands for the summer well mixed open-water period, suggesting the presence of up to eight potentially dominant bacterioplankton groups. The clones derived from DGGE products were dominated by OTUs belonging to uncultured bacteria, including representatives of the Bacteroidetes (Flavobacteria/Sphingobacteria), the β-Proteobacteria and the Actinobacteria. The specific clones identified were the Actinobacteria Corynebacterium glutamicum ATCC 13032, an uncultured Actinobacterium isolate, WL5-10 and an uncultured Crater Lake bacterium CL120-71. Within the Bacteroidetes, specific clones were Flavobacterium xinjiangensis As1.2748, Haliscomenobacter sp. clone SBRT303 and an uncultured bacterium clone SG2-77. Within the β-Proteobacteria, an uncultured bacterium clone SCB5 and clone PRD01a006B were identified (Table 2).

View this table:

Identification of cloned and sequenced DGGE bands from a depth of 5 m in Sombre Lake

DGGE band; sequence A.N.Moss Lake DGGE band closest phylogenetic affiliationStrain/cloneA.N.L. (bp)S. (%)Class/phylumClosest phylogenetic affiliation with a named genus/source
AJ548779C. glutamicumATCC 13032AP00528222100ActinobacteriaCorynebacterium sp.
AJ548780Uncultured Crater Lake bacteriumCL500-95AF31666515998ActinobacteriaUncultured Actinobacterium
AJ548781Uncultured β-ProteobacteriumPRD01a006BAF28915416898β-ProteobacteriaUncultured Polynucleobacter sp.
AJ548782Uncultured bacteriumSG2-77AY1359114597BacteroidetesFlavobacterium frigidarium
AJ548783F. xinjiangensisAs1.2748AF43317315896BacteroidetesFlavobacterium sp.
AJ548784Uncultured Haliscomenobacter sp.SBRT303AF36819011595BacteroidetesHaliscomenobacter sp.
AJ548785Uncultured bacteriumSCB5AF39263519593β-ProteobacteriaUncultured Methylophilus sp.
AJ548786Uncultured Crater Lake bacteriumCL120-71AF31670016693CyanobacteriaUncultured Cyanothece sp.
  • Abbreviations: L. – fragment length in base pairs; S. – similarity in %; A.N. – accession number.

4 Discussion

Relatively few of the genera found were identified independently using at least one culture-dependent and one culture-independent technique (a Sørensen coefficient of 0.17 was calculated between the two groups). The bacterioplankton identified using both approaches were the Actinobacterium group: coryneform bacteria, the Bacteroidetes genera Flavobacterium and Cytophaga, the β-Proteobacterium genus Janthinobacterium and the γ-Proteobacterium genus Pseudomonas. Such results are common to studies adopting a polyphasic approach. Dees and Ghiorse [21] found only one strain, which was identified by both cultivation-dependent and cultivation-independent rDNA cloning methods from 63 isolates and 70 clones. In their study, the hypothesis that easily cultured bacteria are not representative of most bacteria in natural bacterioplankton communities was supported by the large difference between community compositions determined by culture-dependent and culture-independent approaches.

4.1 Preliminary analysis of the bacterioplankton

Earlier culture-based studies showed that Flavobacterium and Cytophaga were dominant in surface waters of Sombre Lake, with a pink Pseudomonad also important in the bottom waters of more eutrophic lakes on Signy Island. Orange colonies dominated surface waters whilst orange and yellow spreading colonies dominated deeper waters. In common with earlier studies, where ectoenzyme activity was detected in Sombre Lake water [51], proteolytic activity was found to be present in over 60% of isolates, though it was particularly significant for white colonies. It was evident that different microbes came to dominance on plates at different incubation temperatures. Thus, at 10°C (a temperature not encountered in the lake environment), Serratia colonies were present in significant numbers, but were virtually absent at 3°C. These results suggest the existence of a propagule bank of potentially significant species, which might be active in the lake environment, but not in culture.

4.2 Direct 16S rDNA sequencing

Most isolates cultured directly from Sombre Lake were typical of freshwater lake bacteria found in a comprehensive study of a temperate freshwater eutrophic lake [52]. These were the β-Proteobacteria genera Janthinobacterium, Pseudomonas and Herbaspirillum, the α-Proteobacteria genus Sphingomonas, the γ-Proteobacteria genus Aeromonas, the Actinobacterium genus Arthrobacter and the Firmicute genera Paenibacillus and Bacillus. Those found in Sombre Lake, but not in the above study, were the β-Proteobacteria genera Aquaspirillum, Duganella, Variovorax and a previously uncultured β-Proteobacterium along with the Firmicute genera Desulfosporosinus and Sporosarcina.

4.3 FAME analysis

Two-thirds of the genera identified from Sombre Lake bacterioplankton by FAME analysis also occurred in the bacterioplankton of the temperate eutrophic lake. Those that did not were the Actinobacteria genera Arcanobacterium, Cellulomonas and Micrococcus, and the γ-Proteobacterium genera Flavimonas and Chryseomonas.

4.4 Clone sequences of 16s rRNA genes

The phylogenetic diversity represented by the cultured bacteria differed significantly from that of the 16S rDNA clone library. PCR-amplified clones represented at least 19 distinct genera. The Actinobacteria and Bacteroidetes (incorporating the CytophagaFlavobacterium group) were particularly common. Interestingly, it has been noted elsewhere that Flavobacterium sp. appear in an increasing number of Antarctic studies [6,53]. Bowman et al. [54] detected members of the Actinobacteria, Verrucomicrobiales, Spirochaetales, Cyanobacteria, low G+C Gram-positives, α-Proteobacteria, β-Proteobacteria, γ-Proteobacteria and Cytophagales in a clone library from three limnologically disparate hypersaline Antarctic lakes. Each of these groups was also represented in Sombre Lake.

4.5 FISH

Data in Fig. 3 suggest that the bacterioplankton community structures of oligotrophic Antarctic lakes are similar to those of temperate regions, being dominated by the β-Proteobacteria followed by the CytophagaFlavobacterium group and the α-Proteobacteria with very low levels of γ-Proteobacteria. As early as 1993, Wagner et al. [28] found that the β-Proteobacteria were more abundant than the γ-Proteobacteria in many aquatic samples. Recent studies using FISH indicate that the β-subdivision of the Proteobacteria seems to be globally distributed and particularly abundant in freshwater environments, along with the next most abundant, the CytophagaFlavobacterium group. In contrast, the CytophagaFlavobacterium group are more abundant than the Proteobacteria in marine ecosystems.

4.6 Clone sequences of DGGE products

The predominant bacterioplankton isolated from DGGE products in our study represented uncultured clones (Table 2). If a certain 16S rRNA gene dominates a clone library it does not imply that it represents a dominant microorganism within the community, unless it is a clone from a DGGE band. DGGE can potentially selectively amplify both the abundant and the more active members of the community, as it is a competitive reaction based on the rRNA gene. Independent evidence, such as in situ hybridisations, would be required to confirm microbial dominance from 16S clone abundance [12,13]. A comparison of DGGE products with the FISH results suggests a potential reason for the discrepancy between EUB 338 and group-specific counts, as two of the DGGE products represent Actinobacteria. The dominance of the β-Proteobacteria and the Bacteroidetes is in agreement with the results from other techniques.

In a recently published study, Zwart et al. [9] found that of 689 bacterial sequences obtained from the water column of rivers and lakes in North America, Europe and Asia, the majority were most closely related to other freshwater clones or isolates. They discerned 34 phylogenetic clusters of closely related sequences that were either restricted to freshwater, or dominated by freshwater sequences. Of these clusters, 23 contained no cultivated organisms. Their putative freshwater clusters were among the α-, β-, and γ-Proteobacteria, the CytophagaFlavobacteriumBacteroides (CFB) group, the Cyanobacteria, the Actinobacteria, the Verrucomicrobia, the green non-sulphur bacteria and candidate division OP10. In an earlier comparison of freshwater studies, Zwart et al. [55] had observed remarkable similarities among 16S rDNA sequences recovered from three different lakes. The major bacterial divisions represented from most of their freshwater sites were the Proteobacteria (α- and β-subdivisions), the CFB group, the Actinobacteria and the Verrucomicrobia. Except for the Verrucomicrobia, the widespread occurrence of organisms from these divisions in freshwater is in agreement with the findings of other in situ hybridisation studies, which used probes for these groups (Glöckner et al. [56]). As was first pointed out by Methé et al. [57], the ubiquity of Proteobacteria of the β-subdivision in freshwater is in sharp contrast to the relative abundance of this subdivision in the open oceans. Of the 34 clusters described, 11 contained cultivated species. The finding that 23 out of 34 clusters contain no cultivated species demonstrates the utility of PCR-based techniques in elucidating microbial diversity in freshwater. In addition, Methé et al. found that the Actinobacteria, the Verrucomicrobia and the CFB group from freshwater contributed many taxa to the clusters, but so far no bacteria have been cultivated in any of these clusters. These groups were also detected in Sombre Lake.

The limitations associated with each of the different techniques used are well documented elsewhere. As discussed earlier, it is well established that growth media exert a dramatic selectivity on the recovery of cultured isolates. Moreover, biases associated with the use of molecular techniques, which may distort the abundance of identified bacteria in a given habitat, are also well known [12,13,16,31,33,55,5860], and are exemplified in the differences between the FAME and 16S rDNA results. Stackenbrandt et al. [61], in a study of an acidic Australian soil, recovered more than 50 isolates of Streptomyces sp. However, none of the 113 16S rDNA clones derived from bulk DNA matched any of the 16S rDNA sequences from the isolates. In fact, only two of the 113 clones were related to Streptomyces sequences even though a Streptomyces-specific primer was used for PCR from soil DNA.

Overall, results from both clone libraries and FISH indicate that the β-Proteobacteria dominate freshwater bacterioplankton [62]. Ward et al. [14] have reviewed several studies in which DNA sequence information from 16S rDNA clones has been compared to the 16S rDNA sequences of the isolates. In extreme environments, it appears that 16S rDNA clones are reasonably representative of the total microbial diversity, due in part to relatively less complex microbial community structure in such environments [12,13]. Prokaryotes dominate many Antarctic ecosystems, and whether or not Antarctic environments contain prokaryotes that are exclusively Antarctic remains an open question [63].

4.7 Summary

From 200 isolates and 100 clones, 42 bacterial genera from six bacterial divisions were identified. Analysis showed that at least six different bacterial divisions, including 24 genera, were identified using culture-dependent techniques, and eight different bacterial divisions including 23 genera were identified using culture-independent techniques (Table 3). Only five genera could be identified using both sets of techniques, however, these five genera came from five different bacterial divisions. Pseudomonas sp., Janthinobacterium lividium (a psychrophilic species), Cytophaga sp., Flavobacterium sp. and coryneform bacteria were all detected by both culture-dependent and culture-independent techniques (Table 4). Important properties that these groups have in common include the ability to utilise a wide variety of organic compounds including complex polysaccharides, are involved in the mineralisation of organic matter, do not require growth factors, are all aerobic (or facultatively anaerobic), their metabolism is mainly respiratory (but in some cases fermentative), and are widely distributed in freshwaters – each of which are properties normally associated with oligotrophic aerobic pelagic environments. However, unlike many of the studies of high latitude lake systems conducted to date, mainly in lakes with a relatively high dissolved organic carbon content (suggesting that they receive input from and are catchment driven), Sombre Lake is controlled by nitrogen and phosphorus availability. Janthinobacterium sp., Flavobacterium sp. and Pseudomonas sp. each contain species that are closely associated with the nitrogen cycle. In addition, the ecologically interesting feature of this particular group of organisms for an Antarctic system is that they all contain species that produce pigment.

View this table:

Groups of bacteria in Sombre Lake identified by culture-dependent and culture-independent techniques

Culture16S sequenceFAME16S cloneDGGE product
View this table:

Sombre Lake bacterioplankton genera identified by technique

Culture16S sequenceFAME16S cloneDGGE product
Embedded Imagexx
Embedded Imagexx
Embedded Imagexxx
Embedded Imagexxx
Embedded Imagexxxx
  • Underlined genera were identified by both culture-dependent and culture-independent techniques.

16S rDNA cloning appears to be as valid as plate cultivation for investigating diversity in the natural environment, so the combination of cultivation-dependent and -independent approaches used in this study yielded complementary information about the composition of the microbial community. In this way, the results reported in this paper expand the knowledge of the biodiversity of Antarctic lake bacterioplankton communities.


This research was supported by the Natural Environment Research Council through the British Antarctic Survey as part of the Terrestrial and Freshwater Biodiversity Project within the Antarctic Biodiversity, Past, Present and Future ASGC Programme. FAME analysis was supported by the Natural Environment Research Council through the Centre for Ecology and Hydrology, Oxford, UK. We would also like to thank Dr Mary L. Edwards (CEH Oxford) for providing preliminary FAME analyses and Drs Alex Rogers and Kevin Newsham (BAS) who commented on the manuscript.


  1. [1]
  2. [2]
  3. [3]
  4. [4]
  5. [5]
  6. [6]
  7. [7]
  8. [8]
  9. [9]
  10. [10]
  11. [11]
  12. [12]
  13. [13]
  14. [14]
  15. [15]
  16. [16]
  17. [17]
  18. [18]
  19. [19]
  20. [20]
  21. [21]
  22. [22]
  23. [23]
  24. [24]
  25. [25]
  26. [26]
  27. [27]
  28. [28]
  29. [29]
  30. [30]
  31. [31]
  32. [32]
  33. [33]
  34. [34]
  35. [35]
  36. [36]
  37. [37]
  38. [38]
  39. [39]
  40. [40]
  41. [41]
  42. [42]
  43. [43]
  44. [44]
  45. [45]
  46. [46]
  47. [47]
  48. [48]
  49. [49]
  50. [50]
  51. [51]
  52. [52]
  53. [53]
  54. [54]
  55. [55]
  56. [56]
  57. [57]
  58. [58]
  59. [59]
  60. [60]
  61. [61]
  62. [62]
  63. [63]
  64. [64]
View Abstract