Comparison of microbial communities associated with phase-separation-induced hydrothermal fluids at the Yonaguni Knoll IV hydrothermal field, the Southern Okinawa Trough

Authors

  • Takuro Nunoura,

    1. Subground Animalcule Retrieval (SUGAR) Program, Extremobiosphere Research Center, Japan Agency for Marine-Earth Science & Technology (JAMSTEC), Yokosuka, Japan
    Search for more papers by this author
  • Ken Takai

    1. Subground Animalcule Retrieval (SUGAR) Program, Extremobiosphere Research Center, Japan Agency for Marine-Earth Science & Technology (JAMSTEC), Yokosuka, Japan
    Search for more papers by this author

  • Editor: Patricia Sobecky

Correspondence: Takuro Nunoura, Subground Animalcule Retrieval (SUGAR) Program, Extremobiosphere Research Center, Japan Agency for Marine-Earth Science & Technology (JAMSTEC), 2-15 Natsushima-cho, Yokosuka 237-0061, Japan. Tel.: +81 46 867 9707; fax: +81 46 867 9715; e-mail: takuron@jamstec.go.jp

Abstract

Microbial communities associated with a variety of hydrothermal emissions at the Yonaguni Knoll IV hydrothermal field, the southernmost Okinawa Trough, were analyzed by culture-dependent and -independent techniques. In this hydrothermal field, dozens of vent sites hosting physically and chemically distinct hydrothermal fluids were observed. Variability in the gas content and formation in the hydrothermal fluids was observed and could be controlled by the potential subseafloor phase-separation and -partition processes. The hydrogen concentration in the hydrothermal fluids was also variable (0.8–3.6 mmol kg−1) among the chimney sites, but was unusually high as compared with those in other Okinawa Trough hydrothermal fields. Despite the physical and chemical variabilities of the hydrothermal fluids, the microbial communities were relatively similar among the habitats. Based on both culture-dependent and -independent analyses of the microbial community structures, members of Thermococcales, Methanococcales and Desulfurococcales likely represent the predominant archaeal components, while members of Nautiliaceae and Thioreductoraceae are considered to dominate the bacterial population. Most of the abundant microbial components appear to be chemolithotrophs sustained by hydrogen oxidation. The relatively consistent microbial communities found in this study could have been because of the sufficient input of hydrogen from the hydrothermal fluids rather than other chemical properties.

Introduction

Recent culture-dependent and -independent microbiological surveys in various deep-sea hydrothermal vent habitats have indicated a variety of community structures and metabolic activities of the indigenous microbial components in the chimney structures and the potential subvent biosphere (Harmsen et al., 1997; Takai & Horikoshi, 1999; Reysenbach et al., 2000a, b; Takai et al., 2001; Schrenk et al., 2003; Nakagawa et al., 2005b; Kormas et al., 2006; Pagéet al., 2008). These studies, coupled with concurrent geochemical and mineralogical characterizations, have led to the presumption of a potential biogeochemical interaction in the habitat and have enabled comparison between the habitats in a hydrothermal field (intrafield comparisons) and even among different fields (interfield comparisons). From the comparisons, it is evident that members of the Thermococcales, Desulfurococcales, Aquificae and Epsilonproteobacteria are cosmopolitan and predominant in the global deep-sea hydrothermal high-temperature environments, and members of Archaeoglobales, Methanococcales, Methanopyrales, Thermodesulfobacteriales, Deferribacteriales and Gammaproteobacteria are also major components in many hydrothermal environments (Takai et al., 2006a; Nakagawa & Takai, 2008).

The emerging patterns of community structures and metabolisms in deep-sea hydrothermal ecosystems could be associated with the physical and chemical conditions of the habitats (Karl, 1995; Takai et al., 2006a). One potential key factor is the concentration of gaseous compounds such as H2, H2S and CH4 in hydrothermal fluids because they serve as the primary energy sources for the chemolithoautotrophs. The abundance and composition of these gaseous compounds in the hydrothermal fluids are substantially controlled by complex processes related to the tectonic, magmatic and hydrogeologic settings. Phase separation and partition of hydrothermal fluid might also have a huge impact on the abundance and composition of the gaseous energy sources (Massoth et al., 1989; Lilley et al., 1993; Butterfield et al., 1994; Nakagawa et al., 2005b; Konno et al., 2006; Takai et al., 2008). Indeed, a phase-separation-associated intrafield variability in functionally active microbial communities has been demonstrated in the hydrothermal fields in the middle Okinawa Trough (Nakagawa et al., 2005b) and in the Lau Basin (Takai et al., 2008). The increased populations of thermophilic and hydrogenotrophic methanogen Methanococcales members (the Iheya North field in the middle Okinawa Trough) and Aquificales members (the Mariner field in the Lau Basin) have been documented as potential responses to the phase-separation-controlled variability in the fluid chemistry.

The Yonaguni Knoll IV hydrothermal field is located at the southwest end of the Okinawa Trough, and dozens of black and clear smoker vent emissions have been identified after its discovery in 2000 (Konno et al., 2006). In sediments extending around hydrothermal vent sites, a large-scale subseafloor liquid CO2 pool originally derived from hydrothermal fluids was also observed (Inagaki et al., 2006; Konno et al., 2006). Multidisciplinary seafloor explorations were conducted in 2003 and 2004 using the manned submersible Shinkai 6500, and many hydrothermal fluid and chimney samples were collected; deployment and recovery of in situ colonization systems (ISCS) (Takai et al., 2003a) in hydrothermal fluids were conducted during the explorations. Geochemical analyses of hydrothermal fluids and the chimney mineral compositions clearly demonstrated the variability in hydrothermal fluid chemistry controlled by the subseafloor phase separation and partition (Konno et al., 2006; Suzuki et al., 2008). In addition, unusually high concentrations of hydrogen in hydrothermal fluids ranging from 0.5 to 5.2 mmol kg−1 were observed (Konno et al., 2006). In this study, we present the structural and metabolic compositions of microbial communities in the various habitats influenced by the variation of the hydrothermal fluid chemistry in order to determine the effects of phase separation of hydrothermalism with a high hydrogen concentration.

Materials and methods

Site description, sampling of hydrothermal fluids and chimneys and deployment of an ISCS

The active area of the Yonaguni Knoll IV hydrothermal field was located at the southwestern end of the Okinawa Trough, where the Ryukyu Arc intersects with the Taiwan Arc (24°50′–24°51′N, 122°51.5′–122°52.5′E). Various hydrothermal fluids and chimney structures were obtained from the ‘Tiger Chimney Mound’, having both black smoker vents (BTC) and clear smoker vents (CTC) (24°50.885′N, 122°42.014′E), the ‘Lion Chimney’, having black smoker vents (LC) (24°50.938′N, 42°020′E), and the ‘Swallow Chimney’, with clear smoker vents (SC) (24°50.832′N, 122°42.013′13E), the manned submersible Shinkai 6500 during the cruises YK03-05 (July 2003) and YK04-05 (May 2004) of the R/V Yokosuka. The self-temperature-recording in situ colonization systems (STR-ISCS) described previously (Takai et al., 2003a) were deployed in the hydrothermal fluid conduits of BTC and CTC vents for 5 and 7 days, respectively, during the YK04-05 cruise. The hydrothermal fluid samples were collected using both a gas-tight fluid sampler ‘Water hydrothermal-fluid Atsuryoku tight sampler II (WHATS II)’ equipped with a self-recording thermometer (Saegusa et al., 2006) and a plastic bag sampler equipped with a deep-sea impeller pump. The samples used in this study are summarized in Table 1. At least two portions of the chimney structures were taken, except for the Swallow chimney site.

Table 1.   Description of the samples used in this study
Sample
ID
Sampling
year
Vent site
(temperature
of vent fluids)
Description of samplesCell density
(cells g−1 or
mL−1)
Archaeal 16S
rRNA gene
amplification
Bacterial 16S
rRNA gene
amplification
Archaeal
population
(%)
Cultivation
analysis
  1. ND, not determined; +, positive; −, negative.

BTC-VE (03)2003Black smoker chimney on Tiger chimney mound (330°C)330°C of vent emission3.3 × 103ND 
BTC-I (03)2003Subsample obtained from the conduit surface of BTC structure2.8 × 105ND 
BTC-S (03)2003Subsample obtained from the outer surface layer (1–3 mm) of BTC structure3.2 × 105++7.2 
BTC-VE (04)2004330°C of vent emission1.0 × 103+NDND
BTC-ISCS (04)2004Deployed in the BTC emission for 5 days1.2 × 104+ND 
BTC-I (04)2004Subsample obtained from the conduit surface of BTC structure5 × 104+ND 
BTC-S (04)2004Subsample obtained from the outer surface layer (1–3 mm) of BTC structure2.8 × 106++16.7 
CTC-VE (03)2003Clear smoker chimney on Tiger chimney mound (330°C)330°C of vent emission3.4 × 103ND 
CTC-I (03)2003Subsample obtained from the conduit surface of CTC structure1.2 × 104ND 
CTC-S (03)2003Subsample obtained from the outer surface layer (1–3 mm) of CTC structure4.8 × 105ND 
CTC-VE (04)2004330°C of vent emission3.0 × 103NDND
CTC-ISCS (04)2004Deployed in the CTC emission for 7 days2.0 × 104ND 
CTC-I (04)2004Subsample obtained from the conduit surface of CTC structure1.0 × 105+ND 
CTC-S (04)2004Subsample obtained from the outer surface layer (1–3 mm) of CTC structure1.0 × 106++38.1 
LC-VE2004Lion chimney (330°C)330°C of vent emission5.0 × 103NDND
LC-I (S)2004Subsample obtained from the conduit surface of LC structure (stem of chimney)1.0 × 104+ND 
LC-S (S)2004Subsample obtained from the outer surface layer (1–3 mm) of LC structure (stem of chimney)2.0 × 105+ND 
LC-I (F1)2004Subsample obtained from the conduit surface of LC structure (flange structure)1.0 × 105ND 
LC-S (F1)2004Subsample obtained from the outer surface layer (1–3 mm) of LC structure (flange structure)6.0 × 105++30.0 
LC-S (F2)2004Subsample obtained from the outer surface layer (1–3 mm) of LC structure (flange structure)ND++1.5ND
SC-I2004Swallow chimney (280°C)Subsample obtained from the conduit surface of SC structure1.0 × 105ND 
SC-S12004Subsample obtained from the outer surface layer (1–3 mm) of SC structure4.2 × 107++7.5 
SC-S22004Subsample obtained from the outer surface layer (1–3 mm) of SC structureND++2.3ND

The chimney structures collected were divided into surface layers and interior structures as described previously (Takai et al., 2001). The synthetic pumice stuffed in ISCS and subsampled chimney structures were divided into three portions. Subsamples for DNA extraction were stored at −80 °C, those for cultivation were anaerobically stored in glass bottles under 100% N2 (200 kPa) with or without 0.05% neutralized Na2S sealed with butyl rubber stoppers and those for cell counting were stored at −80 °C after fixation by filtered (0.22 μm) seawater with 3% formaldehyde at 5 °C overnight. The vent emissions in WHATS gas-tight bottles (c. 100 mL) or plastic bags (5–10 L) were filtered by 0.22-μm pore size cellulose acetate filters and stored at −80 °C for DNA extraction. Ten milliliters of vent emissions were fixed with formaldehyde (final concentration, 3%) and stored at −80 °C for direct cell counting.

Geochemistry of hydrothermal fluids

The geochemistry of hydrothermal fluids was described by Konno et al. (2006) and Suzuki et al. (2008). Several key features are shown in Table 2.

Table 2.   Estimated chemical compositions of end-member hydrothermal fluids, assuming [Mg2+]=0 in the fluid (Konno et al., 2006)
VentTemperature (°C)H2 (mmol kg−1)CH4 (mmol kg−1)CO2 (mmol kg−1)Cl (mmol kg−1)
BTC3300.81.872629
CTC3302.4–3.69.5–13.5306–329332–384
LC3301.0–1.11.2–1.622–47576–674
SC2801.0–1.86.9–7.299–109453–462

Microscopic observation

The chimney suspensions and hydrothermal fluids fixed with 3% formaldehyde were stained with 4′,6′-diamidino-2-phenylindole dihydrochloride (DAPI) (Porter & Feig, 1980). After staining for 1 h at 4 °C, chimney suspensions were centrifuged for 5 s at 2000 g. The supernatants of the chimney suspensions and vent emissions were filtered through a 0.2-μm Isopore membrane filter (Millipore, Ireland). Then, the cells on the filters were counted under fluorescence using an Olympus BX51 microscope. At least 50 microscopic fields for each sample were examined to determine the microbial community density.

Nucleic acid extraction

The DNA assemblage was extracted from chimney structures, ISCS substrata (pumice) and filtered microbial cells in hydrothermal fluids using either the Ultra Clean Soil DNA Purification Mega Kit or the Ultra Clean Soil DNA Purification Kit (Mo Bio Laboratories Inc., Solana Beach, CA). The extracted DNA was further purified by MagExtractor-PCR and Gel Clean up (TOYOBO, Osaka, Japan) if necessary. The DNA of isolates from the cultivation test was also extracted using the Ultra Clean Microbial DNA Kit (Mo Bio Laboratories).

Amplification of archaeal and bacterial 16S rRNA gene and clone analysis

The archaeal and bacterial 16S rRNA genes were amplified from extracted DNA by PCR using LA Taq polymerase with GC buffer (Takara Bio, Otsu, Japan). The oligonucleotide primers for PCR amplification were Arch21F (TTCCGGTTGATCCYGCCGGA) and Arch958R (YCCGGCGTTGAMTCCAATT) or U1492R (ASGGNTACCTTGTTACGACTT) for archaeal 16S rRNA gene, and B27F (AGAGTTTGATCCTGGCTCAG) and U1492R for the bacterial 16S rRNA gene (Lane, 1991; DeLong, 1992). PCR amplification was performed by a thermal cycler GeneAmp 9600 (PE Applied Biosystems, Foster City, CA), and the DNA amplification conditions were 30–40 cycles of 96 °C for 25 s, 50 °C for 45 s and 72 °C for 120 s for the archaeal 16S rRNA gene and 20–30 cycles of 96 °C for 25 s, 54 °C for 45 s and 72 °C for 120 s for the bacterial 16S rRNA gene. The PCR cycle numbers represent the minimum cycle numbers required to provide sufficient amplified products for the cloning based on the preliminary PCR amplification experiments using the same templates.

The amplified gene fragments of the 16S rRNA gene were cloned into pCRII vector (Invitrogen, Carlsbad, CA), and then clone libraries were constructed. The inserts were directly sequenced by the dideoxynucleotide chain-termination method using a dRhodamine sequencing kit (PE Applied Biosystems) according to the manufacturer's recommendations. The primers Arch21F and Bac27F were used for the initial single-strand sequencing of the archaeal and the bacterial rRNA gene, respectively.

The sequence similarity among all of the single-strand gene sequences of the 16S rRNA gene, c. 0.5–0.7 kb long, was analyzed by the fasta-composing algorithm run by dnasis software (Hitachi Software, Tokyo, Japan). The sequences in the clone analysis showing ≥97% and 96% identities of the archaeal and the bacterial 16S rRNA gene, respectively, on dnasis analysis were assigned to the same phylogenetic clone type (phylotype). The sequences of the 16S rRNA gene from isolates showing ≥97% identity on dnasis analysis were tentatively assigned to the same species. The representative rRNA gene clones from each clone type and 16S rRNA gene from each species were further subjected to sequencing, and c. 0.8–1.0 kb of sequences were determined from both strands. Both archaeal and bacterial 16S rRNA gene clone libraries were compared by principal component analysis (PCA) and Jackknife Environment Clusters analysis in unifrac program (http://bmf2.colorado.edu/unifrac/index.psp).

Quantification of the 16S rRNA gene

Quantification of the archaeal and all prokaryotic 16S rRNA genes in the whole microbial DNA assemblages was performed by a quantitative fluorescent PCR method using the 7500 Real Time PCR System (Applied Biosystems) as described previously (Takai & Horikoshi, 2000). The copy number of the 16S rRNA gene in each sample was determined by the average of the triplicate analyses.

Nucleotide sequence accession numbers

The 16S rRNA gene sequences of cultured and uncultured organisms determined in this study were deposited at the DDBJ/EMBL/GenBank nucleotide sequence databases under the following accession numbers: AB235309AB235325, AB235326AB235395 and AB464784AB464836, respectively.

Cultivation analysis

The abundance of viable microorganisms represented by a variety of physiological and metabolic characteristics was determined by a series of serial dilution cultures using each of the hydrothermal fluids and chimney subsamples under the various cultivation conditions. For extremely thermophilic to hyperthermophilic fermentative sulfur-reducing heterotrophs, MJYPS medium (Takai et al., 2000) was used at 70, 85 and 95 °C; for extremely thermophilic to hyperthermophilic methanogens, MMJ medium (Takai et al., 2002) was used at 70, 85 and 95 °C; for thermophilic to hyperthermophilic sulfate reducers, MMJSO medium (Nunoura et al., 2007b) was used at 55, 70, 85 and 95 °C; for mesophilic to hyperthermophilic, strictly anaerobic and autotrophic sulfur reducers, MMJS medium (Nunoura et al., 2008b) was used at 37, 55, 70, 85 and 95 °C; for mesophilic to extremely thermophilic, anaerobic to microaerophilic autotrophs (nitrate-reducing and microaerophilic hydrogen oxidizers and sulfur oxidizers), MMJHS medium (Takai et al., 2003a) with three types of head space gases of 80% H2 and 20% CO2 (2 atm), 79% H2, 20% CO2 and 1% O2 (2 atm) and 75% H2, 20% CO2 and 5% O2 (2 atm) was used at 37, 55 and 70 °C; and for strictly anaerobic thermophilic mixotrophs, MMJYPS medium (Nunoura et al., 2007a) was used at 55 °C. The microorganisms present in the most diluted series of the medium at each temperature were isolated by the subsequent extinction–dilution method (Takai et al., 2000). PCA and cluster analysis of viable populations were conducted using the black-box program (http://aoki2.si.gunma-u.ac.jp/BlackBox/BlackBox.html).

Results and discussion

Variability in hydrothermal fluid chemistry

As reported previously (Konno et al., 2006; Suzuki et al., 2008), the chemical composition of the hydrothermal fluids in the Yonaguni Knoll IV field is variable among the chimney sites (Table 2). The chlorinity of the potential end-member hydrothermal fluids ranges between 332 and 674 mmol kg−1. In contrast to the chlorinity of the fluid, and the concentrations of gas species such as H2, CH4 and CO2 are increased in the Cl-depleted hydrothermal fluids (Table 2). The variability in the chlorinity and the gas content could be associated with the subseafloor phase-separation and -partition processes (Konno et al., 2006). Among the chimney sites studied, the black smoker fluids are Cl-enriched (gas-depleted) and the clear smoker fluids are Cl-depleted (gas-enriched) (Table 2). Interestingly, the Tiger chimney mound has both black (e.g. BTC) and clear (e.g. CTC) smoker vents in an area several meters in diameter, while the black and clear smoker fluids in the Tiger chimney mound show a clear differentiation into Cl-depleted and -enriched fluids (Table 2).

Culture-independent analyses

The compositions of both archaeal and bacterial 16S rRNA gene phylotypes among four different chimney sites are shown in Tables 3 and 4, respectively, and the phylogenetic positions of the archaeal and bacterial phylotypes are indicated in Figs 1 and 2, respectively. The results of 16S rRNA gene amplification are presented in Table 1. Except for the Tiger clear chimney, we could construct two 16S rRNA gene clone libraries from each chimney surface structure.

Table 3.   Distribution of representative archaeal 16S rRNA gene phylotypes in the vent fluid, ISCSs and chimney structures
 Cl-enriched vent fluidsCl-depleted vent fluids
BTC-VE
(04)
BTC-ISCS
(04)
BTC-I
(04)
BTC-S
(03)
BTC-S
(04)
LC-I
(S)
LC-S
(S)
LC-S
(F1)
LC-S
(F2)
CTC- ISCS
(04)
CTC-I
(04)
CTC-S
(04)
SC-S1SC-S2
Crenarchaeota
Desulfurococcales
 pYK04-7A-34  1 1         
 pYK03-5A-3N   14          
 pYK04-8A-2    1         
 pYK04-8A-10    1      31 
 pYK04-14A-15       2  1   
 pYK04-10A-13           1  
 pYK04-10A-26           111
 pYK04-18A-1            1 
 pYK04-18A-2            1 
 pYK04-18A-12            1 
 pYK04-18A-15            1 
 pYK04-18A-31            1 
Thermoproteales
 pYK04-18A-19            1 
 MCGI              
 pYK03-3A-4    2         
 pYK03-3A-15    1         
 pYK04-8A-1    1         
DSAG
 pYK04-19A-7        1     
Euryarchaeota
Thermococcales
 pYK04-1A-1N2324   2424       
 pYK03-12A-3  19 18    231214819
 pYK04-19A-19        1     
Methanococcales
 pYK03-5A-5N   7          
 pYK04-8A-3    5  15  631 
 pYK04-18A-21            1 
Methanopyrales
 pYK04-14A-14       73     
Archaeoglobales
 pYK04-19A-30        2     
 ANME II              
 pYK04-19A-43        1     
DHVE group II              
  DHVE3              
 pYK04-18A-26            1 
 pYK04-19A-46        1     
 PYK04-19A-49        1     
  DHVE4              
 pYK04-19A-8        2     
  DHVE8              
 pYK04-10A-1           7  
 pYK04-20A-4             1
DHVE group I
  DHVE1
 pYK04-19A-1        2     
 pYK04-19A-5        1     
 pYK04-19A-20        1     
 pYK04-19A-34        1     
  DHVE2
 pYK04-8A-26           1  
 pYK04-14A-26       1      
 pYK04-18A-7           1  
 pYK04-19A-6        11     
 pYK04-19A-18        1     
 pYK04-19A-37        2     
 pYK04-19A-45        1     
 pYK04-19A-48        1     
 pYK04-20A-7            112
Unclear affiliation
 pYK04-8A-8    1         
 pYK04-19A-39        1     
Total2324202131242425352319292133
Table 4.   Distribution of representative bacterial 16S rRNA gene phylotypes in the chimney surface habitats
 Cl-enriched vent fluidsCl-depleted vent fluids
BTC-S (03)BTC-S (04)LC-S (F1)LC-S (F2)CTC-S (04)SC-S1SC-S2
Desulfurobacteriaceae
 pYK03-3B-32      
Thermodesulfobacteriaceae
 pYK03-5B-471 1 3 1
 pYK04-20B-52      1
Thermosulfidibacteriaceae
 pYK04-20B-66      1
Deltaproteobacteria & relatives
 pYK03-5B-407 4 1  
 pYK03-5B-891      
 pYK03-5B-961      
 pYK04-10B-25  1    
 pYK04-14B-1   31  
 pYK04-18B-7     1 
 pYK04-18B-12     2 
 pYK04-18B-20     1 
 pYK04-19B-33   1   
 pYK04-19B-72   1   
 pYK04-20B-26      3
 pYK04-20B-51      1
Epsilonproteobacteria
Hydrogenimonaceae (Group A)
 pYK03-3B-2101012 7  
 pYK03-5B-3111     
 pYK04-18B-5     2 
 pYK04-19B-56   1   
Thiovulgaceae (Group B)
 pYK03-3B-191      
 pYK03-9B-46 102    
 pYK04-14B-17  1    
 pYK04-18B-6     1 
 pYK04-19B-2   1   
 pYK04-19B-49   3   
 pYK04-19B-64   1   
Thiovulgaceae (Group F)
 pYK03-8B-11      
 pYK03-8B-4   22 2
 pYK04-7B-10 1  321
 pYK04-14B-37   11312
 pYK04-18B-1     5 
 pYK04-18B-2     6 
 pYK04-18B-4     3 
 pYK04-18B-8     1 
 pYK04-18B-13     21
 pYK04-19B-6   1   
 pYK04-19B-38   3   
 pYK04-19B-50   1   
 pYK04-19B-55   1   
Nautiliales (Group D)
 pYK03-3B-6235 2  
 pYK03-3B-1011  2  
 pYK03-3B-11  1    
 pYK04-7B-19 76 6  
 pYK04-14B-6    1  
 pYK04-14B-10    1  
Thioreductoraceae (Group G)
 pYK03-3B-210 7 12  
 pYK03-5B-672 1    
 pYK04-10B-32  1    
Campylobacteraceae
 pYK04-19B-53   1   
 pYK04-20B-58      1
Alphaproteobacteria
 pYK03-5B-21      
 pYK03-5B-144      
 pYK03-5B-441      
 pYK03-5B-651      
 pYK04-19B-5   1   
 pYK04-19B-59   3   
 pYK04-19B-63   1   
Gammaproteobacteria
 pYK03-5B-282      
 pYK04-14B-5  2    
 pYK04-14B-16  1    
 pYK04-19B-16   2   
 pYK04-19B-67   1   
 pYK04-19B-69   6   
 pYK04-19B-71   1   
Bacteroidetes
 pYK03-5B-451      
 pYK03-5B-531      
 pYK04-19B-70   1   
 pYK04-20B-61      1
Planctomycetes
 pYK04-19B-4   2   
Actinobacteria
 pYK04-19B-7   1   
Acidobacteria
 pYK04-20B-64      1
Chloroflexi
 pYK04-20B-34      1
Deferribacteres       
 pYK04-20B-36      1
OP8
 pYK04-20B-60      1
 pYK04-20B-62      1
TM7
 pYK04-19B-68   1   
 pYK04-20B-13      1
Total39333941292931
Figure 1.

 Phylogenetic analysis of 16S rRNA gene sequences of the representative strains and phylotypes of (a) Crenarchaeota and (b) Euryarchaeota based on neighbor-joining method with 529 and 530 homologous positions, respectively. The boldface type indicates the rRNA gene obtained in this study. MCG I, Marine Crenarchaeotic Group I; DHVEG, Deep-sea Hydrothermal Vent Euryarchaeotic Group.

Figure 2.

Figure 2.

 Phylogenetic analysis of 16S rRNA gene sequences of the representative strains and phylotypes of (a) Aquificae, (b) Epsilonproteobacteria, and (c) Alpha- and Gammaproteobacteria and (d) Deltaproteobacteria based on neighbor-joining method with 711, 603, 603 and 603 homologous positions, respectively. The boldface type indicates the rRNA gene sequences obtained in this study.

Figure 2.

Figure 2.

 Phylogenetic analysis of 16S rRNA gene sequences of the representative strains and phylotypes of (a) Aquificae, (b) Epsilonproteobacteria, and (c) Alpha- and Gammaproteobacteria and (d) Deltaproteobacteria based on neighbor-joining method with 711, 603, 603 and 603 homologous positions, respectively. The boldface type indicates the rRNA gene sequences obtained in this study.

In archaeal 16S rRNA gene clone analysis, the clonal predominance of Thermococcales phylotypes was observed commonly in the Tiger chimney black smoker fluid, the ISCS deployed in both the Tiger black and clear smokers and the internal structures of all the chimneys. Among these samples, only the Tiger black and clear smoker chimney habitats (BTC-I and CTC-I) hosted potentially thermophilic chemolithoautotrophic phylotypes of the Desulfurococcales and/or Methanococcales other than the Thermococcales phylotypes. On the other hand, in the chimney surface habitats, the abundance of Thermococcales was relatively less compared with that in the habitats directly associated with the high temperatures of hydrothermal fluids (Table 3). Instead, potential thermophilic chemolithoautotrophic archaeal phylotypes such as the Methanococcales, Methanopyrales and Desulfurococcales, and a phylotype belonged to previously uncultured Deep-sea Hydrothermal Vent Euryarchaeotic Group II (DHVEG-II) subgroup 8 (Nercessian et al., 2003) showed an increase in their populations (Fig. 1, Table 3). The Archaeoglobales clone was only detected on the surface of the Lion chimney [LC-S (F2)] as a minor population by the 16S rRNA gene clone analysis, although the Archaeoglobales phylotypes have always been identified as one of the dominant archaeal phylotypes in the hydrothermal fields around Japan such as the Iheya North field in the Middle Okinawa Trough (Nakagawa et al., 2005b), and the Myojin Knoll and the Suiyo Sea-mount in the Izu-Bonin Arc (Takai & Horikoshi, 1999; Higashi et al., 2004). The DHVEG-II subgroup 8 was found to be one of the major components in the Tiger clear smoker chimney surface (CTC-S), together with the hyperthermophilic lineages of the Thermococcales, Methanococcales and Desulfurococcales (Table 3). As far as we know, this DHVEG-II subgroup 8 has always been detected along with other hyperthermophilic phylotypes such as the Thermococcales, Methanococcales, Methanopyrales and Archaeoglobales in high-temperature habitats such as the 13°N in the East Pacific Rise, the Suiyo Seamount and the Iheya North field (Nercessian et al., 2003; Higashi et al., 2004; Nakagawa et al., 2005b). Thus, although the DHVEG II subgroup 8 members are uncultivated and their physiology is still unclear, the (hyper)thermophily could be predicted as a key physiological trait of the DHVEG II subgroup 8 based on its habitational preference. Only one of the samples from the Lion chimney surface structure [LC-S (F2)] showed that the DHVE subgroups 1 and 2 predominated the archaeal community. However, considering the growth temperature of Aciduliprofundum boonei, the only isolates in the DHVE 2 that ranged from 55 to 75 °C (Reysenbach et al., 2006), the in situ temperature of the sample might be lower than that of other samples and likely influenced the archaeal community. On the basis of the composition and abundance of the dominating phylogenetic groups, the archaeal rRNA gene community structures in the hydrothermal fluid and chimney habitats in the Yonaguni Knoll IV field more closely resemble those in the Central Indian Ridge (CIR) Kairei field (Takai et al., 2004) than those in the Iheya North field (Nakagawa et al., 2005b), which is geographically closer to the Yonaguni Knoll IV field. Nevertheless, it is also evident that the abundance of hyperthermophilic methanogenic phylotypes in the ISCS and chimney inside structures is much less in the Yonaguni Knoll IV field than in the CIR Kairei field (Takai et al., 2004).

In bacterial 16S rRNA gene clone analysis, we could not obtain indigenous bacterial rRNA genes from high-temperature habitats such as hydrothermal fluids, ISCS deployed in the fluids and chimney interior structures. Other than the Swallow chimney surface habitat (SC-S) samples and one of the samples from the Lion chimney surface habitat [LC-S (F2)], generally similar bacterial 16S rRNA gene community structures were obtained from the chimney surface habitats (Table 4). In the surface habitats of the Tiger black and clear smoker chimneys and the Lion black smoker chimney, the most predominant bacterial phylotypes were affiliated with the thermophilic epsilonproteobacterial family Nautiliaceae and the mesophilic family Thioreductoraceae (group D and G Epsilonproteobacteria) (Nakagawa et al., 2005a; Takai et al., 2005). In addition, bacterial phylotypes related to the genera Thermodesulfobacterium and Balnearium, and the deltaproteobacterial genera were commonly detected in the clone libraries (Table 4 and Fig. 2). The rRNA gene clones within the Alphaproteobacteria, Gammaproteobacteria and Bacteroidetes groups were also found as minor fractions in some chimney surface habitats (Table 4 and Fig. 2). By contrast, in the Swallow chimney surface (SC-S), the epsilonproteobacterial phylotypes of the mesophilic family Thiovulgaceae including the genera Sulfurimonas (Group B), Sulfurovam (Group F) and Nitratifractor (Group F) (Inagaki et al., 2003, 2004; Nakagawa et al., 2005d; Campbell et al., 2006; Takai et al., 2006b) were predominant (Table 4, Fig. 2). In one of the samples from the Lion chimney surface [LC-S (F2)], the predominance of Thiovulgaceae and sulfur-oxidizing Gammaproteobacteria was observed (Table 4, Fig. 2). None of the phylotypes related to the Aquificales were found from the chimney habitats of the Yonaguni Knoll IV field (Table 4 and Fig. 2).

The microbial community density was enumerated by DAPI-stained, direct cell counting. As has always been demonstrated in the previous investigations (e.g. Takai et al., 2008), the community density was much larger on the chimney surface than on the inside portion of the same chimney and in the hydrothermal fluid hosted by the chimney (Table 1). On comparing the cell densities among the surface layers of different chimneys, the Swallow chimney (SC-S) had the largest density (4.0 × 107 cells g−1) and the Lion chimney (LC-S) had the smallest density (6.0 × 105 cells g−1) (Table 1). The proportion of archaeal 16S rRNA gene in the all prokaryotic small subunit rRNA gene assemblages was found to be 7.2% and 16.7% in the Tiger black smoker chimney (BTC-S) samples, 38.1% in the Tiger clear smoker chimney (CTC-S) sample, 30.0% and 1.5% in the the Lion black smoker chimney samples, LC-S (F1) and LC-S (F2), respectively, and 7.5% and 2.3% in the Swallow chimney (SC-S) samples (Table 1). These results imply that the archaeal rRNA gene proportion decreases in the Swallow chimney surface habitat that hosts the largest microbial community density. Based on the bacterial rRNA gene community structure in the Swallow chimney surface samples (SC-S) and one of the Lion chimney surface samples (Table 4), mesophilic Thiovulgaceae phylotypes are likely the populations particularly enriched in the Swallow chimney surface and part of the Lion chimney surface. Thus, the small proportion of archaeal rRNA gene could be explained by increasing populations of mesophilic bacterial components. Furthermore, the dominant members in the archaeal community in LC-S (F2) likely grow at lower temperatures compared with Thermococcales species. The uniqueness of bacterial and archaeal community structures in these samples was reflected in PCA (Fig. 3). Consequently, the variation of the microbial community density and the archaeal rRNA gene proportion may be associated with the variation of the temperature in the Yonaguni Knoll IV field.

Figure 3.

unifrac PCA analysis of 14 archaeal (a) and seven bacterial (b) 16S rRNA gene clone libraries. Blue and red squares indicate Cl-enriched and -depleted fluids vent sites, respectively. 1, BTC-VE (04); 2, BTC-ISCS (04); 3, BTC-I (04); 4, BTC-S (03); 5, BTC-S (04); 6, LC-I (S); 7, LC-S (S); 8, LC-S (F1); 9, LC-S (F2); 10, CTC- ISCS (04); 11, CTC-I (04); 12, CTC-S (04); 13, SC-S1; 14, SC-S2.

Cultivation analysis

The samples used for cultivation analysis are summarized in Table 1. We did not use LC-S (F2) and SC-S2 samples for cultivation analysis, although their 16S rRNA gene communities were presented. We could not obtain any successful cultures from all hydrothermal fluid samples, and the ISCS samples deployed in the Tiger clear smoker and the Lion black smoker fluids as well (Tables 5 and 6). These samples represent the typical habitats consistently exposed to high temperatures, although the 16S rRNA gene clone analysis detected a population of the Thermococcales rRNA genes (Table 3). However, from similar samples, such as the ISCS sample deployed in the Tiger black smoker (BTC-ISCS) and the interior samples of the Tiger black and clear smoker chimneys and the Swallow chimney, the Thermococcales members were found to be one of the most predominantly cultivated populations (Tables 5 and 6). From the ISCS sample of the Tiger Black smoker chimney, only the Thermococcus spp. (strains 83-5-2 and 70-4-2) were detected as a viable population, while the Methanocaldococcus sp. (strain 70-8-3) was also obtained from the Swallow chimney internal habitat (Tables 5 and 6). Not only hyperthermophilic archaeal populations but also thermophilic and mesophilic bacterial members such as Persephonella sp. (strain 70-8-1), Lebetimonas spp. (strains 55S12-1 and 55TY-9-4), Nitratiruptor sp. (strain 55-12-4) and Sulfurimonas sp. (strain 37-8) were obtained from the internal habitats of the Tiger black and clear smoker chimneys (Tables 5 and 6). Among these species, the Thermococcus and Lebetimonas members could be derived from the indigenous microbial components potentially inhabiting the chimney inside structures. Considering their obligate anaerobic and thermophilic traits based on a culturability test, it is difficult to assume that these members are simply contaminated from the exterior habitats during the sample recovery and the subsampling. In contrast, the abundance of the Sulfurimonas sp. strain 37-8 in the interior structure of the Tiger clear smoker chimney (Tables 5 and 6) could be attributed to the contamination from the more abundant population in the exterior habitat of the chimney because of their mesophilic and facultatively aerobic features.

Table 5.   The viable population size of Methanococcales, Thermococcales, Aquificales, Thermodesulfobacteriales, Nautiliales and chemolithoautotrophic Epsilonproteobacteria
Sample
category
Viable population (cells g−1 chimney structure or pumice)
MethanococcalesThermococcalesAquificalesThermodesulfobacterialesNautilialesAutotrophic
Epsilonproteobacteria
  1. The viable population was obtained by serial dilution cultivation analysis that conducted for each two samples for one category except for BTC-ISCS, CTC-ISCS and the Swallow chimney site.

BTC-ISCS 4 × 10    
BTC-I 4 × 100–8 × 1000–6 × 100 0–2 × 106 × 100–2 × 10
BTC-S3 × 102–9 × 1022 × 104–4 × 105 0–3 × 1037 × 1033 × 103–2 × 107
CTC-ISCS
CTC-I 0–4 × 100  0–2 × 1037–2 × 103
CTC-S0–3 × 1022 × 103–2 × 104 3 × 1037 × 103–1 × 1053 × 103–7 × 107
LC-I
LC-S (S)0–1 × 101 × 102–8 × 1041 × 104 1 × 1041 × 10–2 × 104
SW-I3 × 102 × 103    
SW-S3 × 101 × 1052 × 102 2 × 1042 × 104
Table 6.   Distributions and viable numbers of representative strains obtained by serial dilution count cultivation analyses in each category of samples
MediumTemperature (°C)Black smoker (Cl-enriched vent fluids)Clear smoker (Cl-depleted vent fluids)
BTC-ISCSBTC-IBTC-SLC-SCTC-ICTC-SSW-ISW-S
  1. Numbers above representative strain names indicate cell numbers g−1 chimney structure or pumice.

MMJYPS95   2.0 × 10   2.6 × 102
   Pyrococcus sp. str. 95-12-1   Pyrococcus sp. str. 95-12-1
838.0 × 106.0 × 1005.0 × 1048.0 × 1042.0 × 102.0 × 1042.0 × 1031.4 × 105
Thermococcus sp. str. 83-5-2Thermococcus sp. str. 70-4-2Thermococcus sp. str. 83-5-2Thermococcus sp. str. 83-5-2Thermococcus sp. str. 83-5-2Thermococcus sp. str. 83-5-2Thermococcus sp. str. 83-5-2Thermococcus sp. str. 83-5-2
708.0 × 10 4.0 × 1052.0 × 102.0 × 1032.0 × 104  
Thermococcus sp. str. 70-4-2 Thermococcus sp. str. 70-4-2Thermococcus sp. str. 70-4-2Thermococcus sp. str. 83-5-2Thermococcus sp. str. 70-4-2  
MMJ83      3.0 × 10 
      Methanocaldococcus sp. str. 70-8-3 
70  9.0 × 1021.0 × 10   3.0 × 10
  Methanocaldococcus sp. str. 70-8-3Methanocaldococcus sp. str. 70-8-3   Methanocaldococcus sp. 70-8-3
MMJS70     1.0 × 103  
     Thermodesulfobacterium sp. str. 70-S-12  
55     2.0 × 104  
     Lebetimonas sp. str. 55S12-1  
MMJHS70  1.0 × 101.0 × 104 2.0 × 10 2.6 × 102
  Persephonella sp. str. 70-8-1Persephonella sp. str. 70-8-1 Balnearium sp. str. 70-12-3 Persephonella sp. str. 70-8-1
55 6.0 × 1001.0 × 1021.0 × 104 2.0 × 104 2.6 × 102
 Persephonella sp. str. 70-8-1Nitratiruptor sp. str. 55-12-4Nitratiruptor sp. str. 55-12-4 Nitratiruptor sp. str. 55-12-1 Lebetimonas sp. str. 55S12-1
37 2.0 × 101.0 × 1071.0 × 1032.0 × 1037.0 × 1062.6 × 1021.7 × 104
 Sulfurimonas sp. str. 37-8Sulfurimonas sp. str. 37-8Nitratiruptor sp. str. 55-12-4Sulfurimonas sp. str. 37-8Sulfurimonas sp. str. 37-8Thiomicrospira sp. str. 37-SI-2Hydrogenimonaceae str. 37-1%-8-3
MMJHS O2 1%70 6.0 × 100 2.0 × 102    
 Persephonella sp. str. 70-8-1 Hydrogenivirga okinawensis LS12-2T    
55 2.0 × 101.0 × 1021.0 × 1033.0 × 104.0 × 102  
 Nitratiruptor sp. str. 55-12-4Nitratiruptor sp. str. 55-12-4Nitratiruptor sp. str. 55-12-4Nitratiruptor sp. str. 55-12-4Lebetimonas sp. str. 55S12-1  
37 2.0 × 109.0 × 1021.0 × 1042.0 × 1037.0 × 106 2.6 × 102
 Sulfurimonas sp. str. 37-8Hydrogenimonaceae str. 37-1-8-3Nitratiruptor sp. str. 55-12-4Sulfurimonas sp. str. 37-8Sulfurimonas sp. str. 37-8 Nitratifractor sp. str. 37-SO-2
MMJYPS55 2.0 × 107.0 × 103 2.0 × 1021.0 × 105 1.7 × 104
 Lebetimonas sp. str. 55S12-1Caminibacter sp. str. 55YT-8-4 Lebetimonas sp. str. 55TY-9-4Lebetimonas sp. str. 55S12-1 Lebetimonas sp. str. 55S12-1

The microbial components from the surface habitats of the chimney showed high culturability and diversity (Tables 5 and 6). The cultivated microbial community structures of the four chimney surface habitats were generally similar to each other (Table 5); the heterotrophic Thermococcales, the hydrogenotrophic methanogen Methanocaldococcus and the hydrogen- and/or sulfur-oxidizing autotrophic Epsilonproteobacteria (Sulfurimonas, Hydrogenimonaceae and Nitratiruptor) species were the dominant microbial components commonly recovered from all the chimney surface habitats (Fig. 3). In the cultivation test for the S0-reducers at 55 °C, the Nautiliales members such as Lebetimonas and Caminibacter species were always obtained from the highest dilution cultures under both heterotrophic (mixotrophic) (MMJYPS medium) and autotrophic (MMJS and MMHJS media) conditions. The 16S rRNA gene sequence analysis of the strains isolated from the autotrophic and heterotrophic (mixotrophic) cultures indicated that both autotrophic and heterotrophic (mixotrophic) strains were phylogenetically related to each other. Therefore, it seems likely that most of the Nautiliales members isolated from the Yonaguni Knoll IV field are able to grow under both autotrophic and heterotrophic (mixotrophic) conditions, although all the previously isolated and reported strains of Nautiliaceae do not utilize organic carbon as both energy and carbon sources (Alain et al., 2002; Miroshnichenko et al., 2002, 2004; Takai et al., 2005; Voordeckers et al., 2005). These potentially mixotrophic Nautiliales members were also predominant in the chimney surface habitats, except for the Lion chimney in the Yonaguni Knoll IV field.

The potentially mixotrophic Nautiliales members and the hydrogen- and/or sulfur-oxidizing autotrophic Aquificales members (Persephonella spp. and Hydrogenivirga) (Nakagawa et al., 2003; Nunoura et al., 2008a) were also present in abundance in the cultivated microbial communities in most of the chimneys (Tables 5 and 6). Meanwhile, considerable populations of the thermophilic H2-oxidizing S0 or SO42−-reducing autotrophs such as Balnearium and Thermodesulfobacterium spp., respectively (Jeanthon et al., 2002; Takai et al., 2003b), were detected only in the Tiger clear smoker chimney surface habitat (Table 6). A population of the Nitratifractor sp. (strain 37-SO-2) within the Thiovulgaceae was obtained as the Group F Epsilonproteobacteria only from the Swallow chimney surface, in which the predominance of the Group F epsilonproteobacterial phylotypes was shown by the 16S rRNA gene clone analysis (Table 4). Because of the relatively less culturability of Group F Epsilonproteobacteria (Sulfurovam and Nitratifractor) compared with the high culturability of Group B (Sulfurimonas) reported in various hydrothermal vent habitats by Nakagawa et al. (2005c), the discrepancy between the culture-dependent and -independent analyses may have substantially arisen due to the different extents of difficulty in cultivation among the groups of Epsilonproteobacteria. Similar situations may occur in the archaeal populations; the Thermococcales population could be detected by cultivation relatively easily, while the Desulfurococcales members were always identified by molecular analyses.

Intrafield and interfield comparison of the microbial communities in the Yonaguni Knoll IV field

Both culture-dependent and -independent analyses showed that several chimney structures in the Yonaguni Knoll IV field hosted functionally active microbial communities potentially consisting of archaeal and bacterial components such as Thermococcales, Methanococcales, Methanopyrales, Desulfurococcales, Aquificales, Thermodesulfobacteriaceae, Desulfurobactericeae, Gammaproteobacteria and Epsilonproteobacteria. The microbial community structures inferred from either culture-dependent or -independent analysis were different in the composition, for instances, the relative abundance of Group F epsilonproteobacterial phylotypes and species in the Swallow chimney site and the active Thermodesulfobacterium and Balnearium populations specifically in the Tiger clear smoker chimney site. In fact, bacterial 16S rRNA gene populations in each chimney samples were relatively diverse (Fig. 3); however, the composition and diversity of the abundant archaeal rRNA gene phylotypes and the cultivated populations were generally similar among all the chimney sites studied, with a few exceptions (Figs 3 and 4).

Figure 4.

 PCA of viable populations in each chimney surface samples. Blue and red squares indicate Cl-enriched and -depleted fluids vent sites, respectively. 1, BTC-S (03); 2, BTC-S (04); 3, LC-S (S); 4, LC-S (F1); 5, CTC-S (03); 6, CTC-S (04); 7, SC-S1.

In the Yonaguni Knoll IV hydrothermal field, it is already known that the hydrothermal fluids are chemically variable among the chimney sites and the variation of the fluid chemistry could be controlled by the relatively shallow subseafloor phase-separation and -partition processes (Konno et al., 2006; Suzuki et al., 2008). In addition, the phase-separation-associated intrafield variability in the microbial community has been studied intensively in the hydrothermal fields in the middle Okinawa Trough (Nakagawa et al., 2005b) and in the Lau Basin (Takai et al., 2008). The thermophilic and hydrogenotrophic methanogen Methanococcales members in the Iheya North field and of hydrogen- and/or sulfur-oxidizing Aquificales members in the Mariner field were particularly abundant in the habitats associated with the Cl-depleted (gas-enriched) hydrothermal fluids in these fields (Nakagawa et al., 2005b; Takai et al., 2008). In both cases, it was suggested that the phase-separation-induced H2 enrichment might be a major factor responsible for the increased populations of hydrogenotrophic chemolithoautotrophs. The potential phase-separation-induced H2 concentration anomaly (0.8–3.6 mmol kg−1) is also observed in the hydrothermal fluids of the Yonaguni Knoll IV field. If the H2 enrichment in the hydrothermal fluids were to have an impact on the hydrogenotrophic components of the microbial communities as in the cases of the Iheya North and the Mariner fields, some of the Methanococcales, Aquificales and Epsilonproteobacteria phylotypes and species should be more abundant in the community in the Tiger clear smoker chimney (the most H2-enriched) than in the Tiger black smoker chimney (the least H2-enriched). Neither culture-dependent nor -independent characterization in this hydrothermal field supports this assumption. Neither 16S rRNA gene community structures nor viable microbial communities show any difference between Cl-enriched or -depleted vent sites in PCA and cluster analysis (Figs 5 and 6), and some of the differences in 16S rRNA gene clone analysis may be explained by the in situ temperature of each sample as described above. The specific viable population of hydrogenotrophic Thermodesulfobacterium and Balnearium species and the most abundant occurrence of the potentially mixotrophic Nautiliales members in the Tiger clear smoker chimney may represent a possible response of the microbial community to the H2 enrichment. However, it is still unclear, without further detailed biogeochemical and microbiological characterizations, whether the increased culturability of these thermophilic hydrogenotrophs is a response to the enriched H2 in the hydrothermal fluid or to other physical and chemical conditions of the habitat. It can be said that the chemical variability in the phase-separation-controlled hydrothermal fluid has much less impact on the composition and function of the microbial communities among the chimney sites in the Yonaguni Knoll IV field than in the Iheya North (Nakagawa et al., 2005b) and Mariner (Takai et al., 2008) fields. The variation of the H2 concentration in the end-member hydrothermal fluids is 45–96 μmol kg−1 and 12–130 μmol kg−1 in the Mariner field (Takai et al., 2008) and the Iheya North field (K. Takai et al., unpublished data), respectively. The variation of H2 in these fields occurs at a one magnitude lower concentration than in the Yonaguni Knoll IV field (0.8–3.6 mmol kg−1). Thus, one possible explanation for the relatively less intrafield variability in the microbial community may be the sufficient supply of H2 and other energy sources to the chimney habitats from the hydrothermal fluids even in the Cl-enriched (gas-depleted) vent sites for the Yonaguni Knoll IV field. Nevertheless, the geological and geochemical mechanisms behind why the hydrothermal fluids in the Yonaguni Knoll IV field are highly enriched with H2 are still unclear, even though the field is located in the typical Ryukyu Arc – Backarc system (Suzuki et al., 2008). Further investigation of geographically and geologically distinct hydrothermal systems and interfield comparisons may possibly provide key insights into the link between geological, physical, chemical and microbiological settings of the deep-sea hydrothermal systems.

Figure 5.

 Cluster analyses of clone libraries. The trees were created with Jackknife Environment Clusters analysis of the unifrac program. Jackknife with 100 permutations was performed. Jackknife values over 50 are given at corresponding branches. Comparisons of archaeal 16S rRNA gene clone libraries from all samples (a) and from chimney surface samples (b), and of bacterial 16S rRNA gene clone libraries (c).

Figure 6.

 Cluster analyses of viable populations in chimney surface samples. Comparisons of viable populations in chimney surface samples (a) and that of maximum viable populations in each vent sites (b).

Acknowledgements

We thank R/V Yokosuka and Shinkai 6500 operation teams during the cruises YK03-05 and YK04-05 (JAMSTEC) for their assistance in collecting samples.

Ancillary