High incidence of halotolerant bacteria in Pacific hydrothermal-vent and pelagic environments


  • Jonathan Z Kaye,

    Corresponding author
    1. University of Washington, School of Oceanography, Box 357940, Seattle, WA 98195-7940, USA
      *Corresponding author. Tel.: +1 (206) 616-9041; Fax: +1 (206) 543-0275 jzkaye@ocean.washington.edu
    Search for more papers by this author
  • John A Baross

    1. University of Washington, School of Oceanography, Box 357940, Seattle, WA 98195-7940, USA
    Search for more papers by this author

*Corresponding author. Tel.: +1 (206) 616-9041; Fax: +1 (206) 543-0275 jzkaye@ocean.washington.edu


The abundance of halotolerant microorganisms in hydrothermal-vent and pelagic waters in the North and South Pacific was estimated by the most probable number (MPN) technique using a heterotrophic 16% NaCl medium incubated at 20–24°C. Based on these MPNs and direct counts with epifluorescence microscopy to enumerate the total microbial population, salt-tolerant microbes comprised from <0.01 to >28% of the total microbial community. Fourteen isolates from these MPN enrichments were identified by sequencing a portion of the 16S rRNA gene, and all were found to belong to the genera Halomonas and Marinobacter. The response to salt of mesophilic hydrothermal-vent microbial isolates obtained without selecting for salt tolerance was also examined. Forty-one of 65 strains cultured from hydrothermal plume waters, low-temperature hydrothermal fluids, sulfide rock and an animal specimen at ∼2000–2200 m depth from the Endeavour Segment of the Juan de Fuca Ridge were subjected to increasing concentrations of NaCl, and over half grew at a NaCl concentration that is lethal to many commonly isolated marine bacteria. At least 36 of the 65 isolates (≥55%) grew in the enrichment medium supplemented with 10% NaCl; at least 30 of 65 (≥46%) grew with 16% NaCl; at least 20 of 65 (≥31%) tolerated 22% NaCl. Based on phylogenetic analysis of the 16S rRNA gene in nine of these 65 isolates, four belonged to the genus Halomonas. These Halomonas strains tolerated 22–27% NaCl. It is possible that a majority of the other 16 isolates which grew with 22% NaCl are also Halomonas based on their degree of halotolerance, morphology, and apparent abundance as revealed by MPN enrichments. The four Halomonas strains obtained without selecting for halotolerance were further characterized physiologically and metabolically. Overall, they grew between −1°C and 40°C, were facultative aerobes, oxidized between 49 and 70 organic compounds according to Biolog plate substrate utilization matrices, grew with oligotrophic quantities of carbon (0.002% yeast extract) in liquid media, reduced nitrate to nitrite, and tolerated up to 0.05–3 mM Cd2+. Halomonas is one of the most abundant culturable organisms in the ocean, and its success may be attributed to its metabolic and physiological versatility.


Molecular phylogenetic analyses of prokaryotic communities in pelagic marine environments reveal that they consist predominantly of clusters of uncultured bacterial and archaeal species [1–3]. For example, the small-subunit rRNA sequence databases show that as yet uncultured bacteria belonging to the α-subclass of the Proteobacteria (e.g. SAR clusters) comprise up to 25% of the total microbial community [3] and are abundant in both Pacific and Atlantic waters [4]. The role that these numerically dominant microorganisms play in oceanic ecosystems remains elusive however, because nothing is known about their metabolism or physiology [3,5] and because the phylogenetic distance is large between these organisms and nearly all characterized bacteria, rendering phenotypic inferences unreliable [5].

An understanding of the metabolic and other relevant ecological functions of these numerically dominant microbes generally necessitates their isolation and characterization using pure-culture techniques. Media designed to enrich for uncultured pelagic microorganisms, which are hypothesized to be dominated by oligotrophs, attempt to mimic in situ conditions by using low concentrations of organic carbon [5]. Efforts to culture such numerically dominant microbes have made slow progress, and still less than 0.1% of the total microbial community can be cultured on an array of heterotrophic solid media [5]. This low incidence of culturability among heterotrophs is attributed to imbalances in or inappropriate mixtures of organic compounds in enrichment media, inhibition by other organisms, and viral lysis due to the improved nutritional conditions provided to the bacteria [5]. However, the marine oligotroph Sphingomonas sp. strain RB2256 was first isolated by dilution culture [6], and it may be a metabolically active and numerically abundant species in some pelagic environments [5,7]. Culturing very high percentages of the total microbial count has been achieved by enumerating aggregate-forming units on solid media with epifluorescence microscopy [8], but this approach has limitations and does not allow for the isolation and subsequent characterization of microbial strains. Overall, attempts to isolate numerically dominant microorganisms most often revolve around creating in situ conditions by adjusting media compositions, culturing surfaces (e.g. agar plates, membrane filters), or dilutions with unamended filtered seawater [5,6,8].

Little is known about the microbial community structure in the deep-sea, mid-ocean ridge setting. The hydrothermal-vent environments associated with mid-ocean ridges include positively buoyant and neutrally buoyant (lateral) hydrothermal plumes (∼2°C), low-temperature hydrothermal fluids (∼5–100°C), high-temperature hydrothermal fluids (∼150–400°C), sulfide rock, basalt, and pelagic and metalliferous sediments [9–12]. Vent waters and bottom seawater in close proximity to hydrothermal vents are typically enriched in inorganic metabolic energy sources, including H2, CH4 and Mn [9,13]. The concentration of organic carbon in hydrothermal plumes is low, similar to the non-vent, deep-sea water column (D. Butterfield, unpublished data). Community phylogenetic analyses have focused on microbial mats and high-temperature hydrothermal fluids [14,15], and culturing work has been biased towards hyperthermophiles, thermophiles, and mesophilic sulfur- and metal-metabolizing chemoautotrophs [10,12,16].

We isolated numerous bacterial strains from hydrothermal plumes, low-temperature hydrothermal fluids, and other vent samples using an oligotrophic medium amended with enriched levels of reduced transition metals and H2 or CH4 in some instances. A significant portion of these isolates were found to tolerate high levels of salt and heavy metals [17]. Here we report on the abundance and characterization of these and other halotolerant microorganisms in both hydrothermal-vent and pelagic environments.

2Materials and methods

2.1Enrichment medium

The isolation medium contained synthetic seawater (Sea Salts B), a trace elements solution (Trace Elements F) and additional nutrients. Sea Salts B has (per liter of deionized water) 19.6 g NaCl, 3.3 g Na2SO4, 0.5 g KCl, 0.05 g KBr, 0.02 g H3BO3 and 8.8 g MgCl2·6H2O [18]. The medium was supplemented with 10 ml of the trace elements solution (per liter of Sea Salts B). Trace Elements F consists of (per liter of deionized water) 0.05 g Al2(SO4)3, 0.1 g H3BO3, 0.05 g LiCl, 0.1 g Na2MoO4·2H2O, 0.05 g KBr, 0.05 g KI, 0.05 g NaF, 0.1 g ZnSO4·7H2O, 0.005 g BaCl2, 0.005 g CoCl2·6H2O, 0.01 g CuSO4·5H2O, 0.2 g MnCl2·4H2O, 0.01 g NiCl2·6H2O, 0.005 g Na2SeO4, 0.005 g SrCl2·6H2O, 0.005 g H2WO4 and 0.005 g VOSO4·xH2O [18,19]. For enrichments in 1991, the trace elements solution contained 0.015 g NiCl2·6H2O but no NaF. Lastly, the additional components of the isolation medium include (per liter of Sea Salts B) 1.605 g NaNO3, 5.0 g Na2S2O3·5H2O, 0.02 g yeast extract, 1.0 g PIPES buffer (piperazine-N,N′-bis[2-ethanesulfonic acid] disodium salt), 0.002 g FeSO4·7H2O, 0.15 g MnSO4·H2O, 0.1 g CaCl2, 0.430 g (NH4)2SO4 and 0.036 g KH2PO4. Trace Elements F, FeSO4·7H2O, MnSO4·H2O, CaCl2, (NH4)2SO4, 0.605 g of the NaNO3 and KH2PO4 were added after autoclaving via filter sterilization. When necessary, the medium was solidified with 13 g (per liter) purified agar (BBL). The pH was adjusted to 7 or 9 with filter-sterilized 1 M HCl or 1 M KOH if needed. Some agar slants were prepared in stoppered serum bottles with a H2/CO2 (80:20 vol.%) or CH4 headspace achieved by purging and filling the bottles with gas four times [20]; others simply contained an air headspace.

The medium used for most probable number (MPN) enrichments contained (per liter of isolation medium) 156.8 g NaCl and 1.0 g sodium citrate as well.

2.2Quantitative enrichments for halotolerant microorganisms

To quantify halotolerant microorganisms in the vent environment and overlying water column, the three-tube MPN technique [21] was used with the growth medium at an elevated NaCl concentration of 16% and with 0.1% sodium citrate. Forty-three samples of low-temperature hydrothermal fluid, buoyant hydrothermal plume, lateral hydrothermal plume, bottom seawater in close proximity to hydrothermal vents, and non-vent seawater from 10 to 2800 m depth were collected with the following samplers: (1) a Niskin bottle rosette with a conductivity–temperature–depth–transmissometry (CTDT) package to detect hydrothermal plume temperature and particle anomalies; (2) Niskin bottles attached to the Deep-Submergence Vehicle Alvin basket; (3) the Remote-Operated Vehicle ROPOS suction sampler; (4) the Hot Fluid Sampler (NOAA-PMEL VENTS Program, Seattle, WA, USA) mounted on ROPOS; or (5) titanium syringe samplers [22] triggered by Alvin. These samples were procured in 1998 from within or above the Mothra vent field on the Endeavour Segment (48°N, Juan de Fuca Ridge (JdFR)), Axial Seamount (46°N, JdFR), and 17.5–21.5°S along the Southern East Pacific Rise (SEPR). Seawater or vent water was diluted with autoclaved Sea Salts B in sterile Falcon tubes. Between 0.02 μl and 200 μl of sample water was then inoculated in triplicate onto agar plates or into broth and incubated at room temperature (20–24°C) onboard ship. Growth of colonies or turbidity in tubes indicated preliminary positive results. A dissecting microscope was used to verify the presence of colonies if necessary, and visualization with phase-contrast microscopy (Zeiss) confirmed all positive and negative results for the MPNs in broth.

Thirteen cultures were isolated from the most dilute positive MPN enrichment tube of a given sample by transferring 100 μl from that MPN tube onto an agar plate of the same composition, and subsequently isolating colonies by triplicate streak transfers. One strain (A-sw1) was isolated from the second-most dilute MPN enrichment tube in the same manner.

2.3Counts of total microorganisms

Each water sample (36–37 ml) was preserved with either glutaraldehyde (type II, 2% final concentration) or formalin (4% final concentration) and then stored at 2°C for 9–14 months before counting. Aliquots of fixed sample were filtered onto 0.22-μm black polycarbonate filters (Osmonics, Inc.) and stained with 4′,6-diamidino-2-phenylindole. Cells were enumerated by epifluorescence microscopy (Zeiss) [23].

2.4Isolation of oligotrophs without selecting for halotolerance

Oligotrophic microbes were initially enriched from hydrothermal-vent samples in 1991 and 1995, all from ∼2000–2200 m depth from the Main Endeavour Field on the Endeavour Segment of the JdFR. Inocula included hydrothermal plume water, low-temperature hydrothermal fluids, sulfide rock and a vestimentiferan (Ridgeia piscesae) trophosome. Hydrothermal plume temperature and particle anomalies were detected with a CTDT package and sampled with a rosette of Niskin bottles. Rocks and animals were procured with Alvin manipulators.

Solid samples were placed directly into the enrichment medium. Water samples (1, 10 and 50 ml) were first filtered onto 0.2-μm polycarbonate filters and then placed onto agar slants in 55-ml serum bottles. Samples were incubated at either 2 or ∼20°C aboard ship.

2.5Phenotypic characterization of vent isolates

Nine oligotrophic isolates were characterized physiologically and metabolically. After isolation, they were grown on agar plates and in broth composed of a medium slightly modified from the enrichment medium. This growth medium always had an air headspace, was made at pH 7, and contained 0.1% sodium citrate to achieve higher cell densities. Unless otherwise noted, all experiments occurred at 23°C, room temperature. Halomonas pacifica and Halomonas aquamarina were obtained from the American Type Culture Collection.

The Gram stain was performed according to a standard procedure [24].

The ability to grow at elevated salt concentrations was assessed with the growth medium augmented with NaCl. Growth was monitored by phase-contrast microscopy (Zeiss) and scored as positive if the concentration of cells reached ≥107 ml−1 from an initial concentration of ≤105 ml−1.

To determine which organic carbon compounds could be metabolized, Gram-negative Biolog plates (Biolog, Inc., Hayward, CA, USA) containing 95 wells pre-filled with different organic substrates and a color-sensitive metabolic indicator were used; each plate also has a control well pre-filled only with the metabolic indicator [25]. Cells were grown to the exponential phase in Falcon tubes to a density of ∼108 cells ml−1 and pelleted by spinning in a clinical centrifuge at maximum speed at room temperature for 20–40 min. The supernatant was decanted, and the pellet was resuspended in 15 ml of the growth medium modified by removing the yeast extract and sodium citrate. 150 μl of this suspension was inoculated into each of the 96 wells of the Biolog plate, which were then incubated at 23°C for at least 4 days in a moisture chamber and scored every 8–24 h. All control wells remained negative, indicating that the residual organic carbon (from the growth medium) retained by the pellets and subsequently inoculated into the wells had a negligible impact on results.

Tolerance to divalent cadmium was assayed by adding Cd to the growth medium from a stock of filter-sterilized 1.52% CdCl2·2.5H2O in Sea Salts B.

For anaerobic growth, the growth medium was dispensed into Balch tubes which were sealed with rubber stoppers and aluminum crimps. The headspace was purged of air and filled with argon four times [20]. The medium was reduced with filter-sterilized Na2S·9H2O (0.05%), and resazurin (0.0002%) indicated oxygen removal. Anaerobic nitrate reduction was assayed colorimetrically via nitrite production [26].

2.6Phylogenetic analyses

Twenty-three strains were identified phylogenetically: 14 isolated from MPN enrichments and nine oligotrophs isolated without selection for halotolerance from various hydrothermal-vent samples. Pure cultures of cells were grown in 500-ml flasks and pelleted by centrifuging at 10 000×g for 20–40 min at 4°C. Genomic DNA was extracted using the IsoQuick kit (Orca Research, Bothell, WA, USA). A portion of the 16S rRNA gene was amplified by polymerase chain reaction (PCR) using the bacterial primers 8F (5′-AGA GTT TGA TCC TGG CTC AG) and 519R (5′-GWA TTA CCG CGG CKG CTG) [27]. After cleaning the PCR products, they were re-amplified with the same primers and fluorescently tagged dideoxyribonucleotides and then sequenced with an Applied Biosystems sequencer (ABI) Model 373A at the University of Washington Marine Molecular Biotechnology Laboratory or an ABI100 at the Molecular Genetics Instrumentation Facility at the University of Georgia, Athens, GA, USA. Sequences of approximately 400 bp length (bases 100–493 and 101–484, Escherichia coli numbering, in Figs. 2 and 3, respectively) were aligned with other sequences acquired from GenBank using the Ribosomal Database Project website [28] after excising a variable stem-loop (bases 198–219, E. coli numbering, in Figs. 2 and 3, and additionally bases 450–483, E. coli numbering, in Fig. 2). Phylogenetic trees were created with PAUP* (D.L. Swofford, Smithsonian Inst.) as found in the Genetics Computer Group (GCG) (Wisconsin Package version 9.1, GCG, Madison, WI, USA) using a neighbor-joining distance algorithm [29,30] with negative branch lengths prohibited and with bootstrap analysis (100 replicates). Trees were visualized with TreeView [31]. GenBank accession numbers are E-plume1 AF212202; E-plume2 AF212201; E-plume3 AF212203; E-sulfide1 AF212204; M-sw1 AF212205; A-lthf1 AF212206; A-plume1 AF212207; A-sw1 AF212208; A-sw2 AF212209; S-plume1 AF212210; S-plume2 AF212211; S-sw1 AF212212; S-plume3 AF212213; S-bplume1 AF212214; S-sw2 AF212215; S-plume4 AF212216; S-lthf1 AF212217; S-lthf2 AF212218; E-limpetgut1 AF251770; E-twt2 AF251771; E-twt1 AF251772; E-lthf1 AF251773; and E-plume4 AF251774.

Figure 2.

Phylogenetic tree of nine hydrothermal-vent isolates obtained without selecting for tolerance to salt. The tree was constructed by aligning ∼400 bp of the 16S rRNA gene with known organisms. Scale bar indicates 10% sequence change. Bootstrap values (100 replicates) given at branch points.

Figure 3.

Phylogenetic tree of 18 Halomonas and Marinobacter strains, constructed as in Fig. 2. Fourteen strains were isolated from the most dilute MPN enrichments (second-most dilute for A-sw1), and four were isolated without selecting for halotolerance. Scale bar indicates 10% sequence change. Bootstrap values (100 replicates) given at branch points.


The MPNs revealed that microorganisms capable of growing with 16% NaCl in hydrothermal-vent and pelagic waters ranged from <15 to >12 000 ml−1 (95% confidence limits: <2.5 to 65 000 ml−1) (Table 1). The total number of microbes in these samples varied from (7.3±2.0)×103 ml−1 to (6.2±1.2)×105 ml−1. Overall, these culturable halotolerant microbes comprised between <0.01 and >28% of the total population. Combining the 95% confidence limits of the MPNs and total counts shows that the absolute minimum percentage is <0.00% and absolute maximum percentage is >100%. There is no apparent relationship between sample type or geographic location and percentage of halotolerant microbes (Fig. 1). Thirteen bacterial strains were isolated from the greatest dilution (and one from the second greatest dilution, A-sw1) of different MPN enrichments and analyzed phylogenetically. Eight belonged to the genus Halomonas and six belonged to the genus Marinobacter.

Table 1.  Range in the number of culturable halotolerant microorganisms (ml−1) based on three-tube MPN estimates using a 16% NaCl heterotrophic medium, range in the total number of cells (ml−1) based on epifluorescence microscopy, and range in the percentage of total microorganisms which are halotolerant, in seawater and hydrothermal-vent samples
  1. aMean of one subsample ±95% confidence limits calculated from n=32–64 fields of a microscopic grid.

  2. bIsolate name followed by generic designation, Marinobacter or Halomonas.

Sample typeDepth (m)Halotolerant microbes (ml−1)Total microbes ×104 (ml−1)aHalotolerant percentage (%)Isolateb
Surface seawater10–3045–4 70026±6.5 to 62±120.01–1.4A-sw2, Mar.
Mid-water500–1 40045–12 0002.0±0.57 to 7.4±1.00.06–26A-sw1, Mar.
     M-sw1, Hlm.
     S-sw2, Hlm.
Deep water2 000–2 780230–5500.72±0.20 to 1.3±0.351.9–7.6S-sw1, Hlm.
Lateral plume1 530–2 61012–5503.4±1.0 to 18±4.30.01–1.6A-plume1, Mar.
     S-plume1, Mar.
     S-plume2, Mar.
     S-plume3, Mar.
     S-plume4, Hlm.
Buoyant plume2 590–2 8305 500–>12 0004.1±1.2 to 17±3.13.3–>28S-bplume1, Hlm.
Bottom seawater1 560–2 700<15–4 7003.2±0.66 to 13±1.8<0.01–12 
Low-temperature fluid1 520–2 820<15–>12 0001.6±0.25 to 35±5.4<0.02–>3.5A-lthf1, Hlm.
     S-lthf1, Hlm.
     S-lthf2, Hlm.
Figure 1.

Halotolerant percentage of the total microbial community. Percentages are derived from dividing three-tube MPN estimates of the number of halotolerant microorganisms which grew on a 16% NaCl, 0.1% sodium citrate medium incubated at 20–24°C by enumerations of the total microbial community obtained with epifluorescence microscopy. Samples include non-vent ocean water, buoyant and lateral hydrothermal plumes, bottom seawater in close proximity to hydrothermal vents, and low-temperature hydrothermal fluids from the North and South Pacific. MPN values which provide percentages that are less than a given number (e.g. <0.01%) are not included; values which provide percentages that are greater than a given value (e.g. >28%) are included.

In 1991 and 1995, 17 samples of hydrothermal plume water, low-temperature hydrothermal fluid, sulfide rock, and a tubeworm trophosome were inoculated onto the enrichment media and incubated at 2°C or ∼20°C (Table 2). Sixty-five microbial strains were then purified from these primary enrichments. Forty-one of these 65 strains remained viable until 1997 and were screened for tolerance to increasing levels of NaCl (Table 2). At least 36 (≥55% of the original 65) grew in the enrichment medium with 10% NaCl; at least 30 (≥46% of 65) grew with 16% NaCl; and at least 20 (≥31% of 65) grew with 22% NaCl. A minimum of five (≥8% of 65) could not grow at a NaCl concentration at or above 10%. The 20 strains which grew with 22% NaCl were each rods, 1 μm by 1.5–4 μm in dimension.

Table 2.  Sample source, incubation conditions and degree of halotolerance among 41 isolates
  1. aIsolate name followed by generic designation, if determined: Halomonas and Pseudoalteromonas abbreviated.

SampleSample sourceIncubationIsolateaSalt tolerance (% NaCl)
 Sample typeT (°C)YearHeadspaceT (°C)pH  
ALateral plume21991CH427E-plume1, Hlm.25
BLateral plume21991CH427E-plume2, Hlm.22
CLateral plume21991H2/CO227E-plume722
DLateral plume21991H2/CO227E-plume822
       E-plume3, Hlm.27
ELateral plume21991H2/CO227E-plume1022
GLateral plume21991H2/CO229E-plume1310
HLateral plume21991H2/CO229E-plume1510
ILateral plume21991H2/CO227E-plume1722
JLateral plume21991H2/CO227E-plume4, new genus?14
KLateral plume21991H2/CO227E-plume2116
LLow-temperature fluid∼101995air∼207E-lthf216
       E-lthf1, new genus?16
MRidgeia trophosome∼101995air∼207E-twt322
       E-twt1, Pseudoalt.16
       E-twt2, Vibrio8
NSulfide rock∼101995air∼207E-sulfide1, Hlm.22
OLateral plume21991CH429E-plume2522
PLateral plume21991CH429E-plume262
QLateral plume21991CH429E-plume2722
RLimpet gut∼101995air∼207E-limpetgut1, Vibrio8

A portion of the 16S rRNA gene was sequenced for nine of the 65 isolates (Fig. 2). These nine organisms were chosen based on differences in morphology, colony appearance, Biolog plate assays and sample source, but not based on salt tolerance. Four of the nine strains (E-plume1, E-plume2, E-plume3 and E-sulfide1) belonged to the genus Halomonas. The remaining five belonged to Pseudoalteromonas or Vibrio, or were distantly related to named genera; isolate E-plume4 was closely related to an unnamed Mn-oxidizing bacterium, however [32].

A phylogenetic tree was constructed with ∼400-bp 16S rRNA sequences from the 18 Halomonas and Marinobacter strains (14 from MPN enrichments and four from oligotrophic enrichments without selecting for halotolerance) and previously characterized Halomonas and Marinobacter species (Fig. 3). Isolates A-plume1, S-plume1, A-sw1, A-sw2, S-plume2 and S-plume3 formed a cluster within the genus Marinobacter. Strains A-lthf1, S-sw1, S-sw2, S-lthf1 and E-plume2 clustered with H. aquamarina and Halomonas meridiana. Halomonas venusta fell out with S-bplume1, S-plume4 and S-lthf2. Strains E-plume1 and E-plume3 formed their own branch within the Halomonas genus, as did Halomonas variabilis and isolate E-sulfide1. Lastly, isolate M-sw1 and Halomonas marina are most closely related to each other. High bootstrap values at the Halomonas and Marinobacter generic branch points (99 and 88, respectively) highlight the robustness of the genus-level groupings.

Halomonas isolates E-plume1, E-plume2, E-plume3 and E-sulfide1, along with type cultures of H. pacifica and H. aquamarina, were further characterized phenotypically (Table 3). The four novel strains were motile, Gram-negative, and grew with 22–27% NaCl. Colonies were tinted orange and had an encrusted appearance. The minimum temperature that permitted growth was −1°C for E-plume1, E-plume3 and E-sulfide1, and 2°C for E-plume2, H. aquamarina and H. pacifica. While H. aquamarina and H. pacifica grew at 45°C, these four new Halomonas strains grew only up to 40°C. Addition of Cd2+ retarded growth, but only at extremely high (mM) levels for three of the Halomonas strains. Strain E-plume2 grew in the presence of 3.0 mM Cd2+, and strains E-plume1 and E-plume3 tolerated 2.0 mM Cd2+. H. pacifica survived with 0.5 mM Cd2+ while isolate E-sulfide1 and H. aquamarina only tolerated 0.05 mM Cd2+.

Table 3.  Temperature range, metabolism and cadmium tolerance of four novel Halomonas strains isolated without selecting for halotolerance, H. aquamarina and H. pacifica
  1. aNot determined.

OrganismTemperature range (°C)Carbon compounds oxidized (Biolog plate) (% of 95 tested)Anaerobic nitrate reductionCd2+ tolerance (mM)
E-plume1, Halomonas−14054+2.0
E-plume2, Halomonas24052+3.0
E-plume3, Halomonas−14062+2.0
E-sulfide1, Halomonas−14074+0.05
H. aquamarina245a+0.05
H. pacifica245480.5

These four Halomonas isolates were heterotrophic facultative aerobes, and each reduced nitrate to nitrite. According to Biolog plates, strains E-plume1, E-plume2, E-plume3 and E-sulfide1 overall oxidized 52–74% of 95 provided organic compounds. All four strains utilized a variety of 42 sugars, starches, amino acids, organic acids and other organic compounds. None of the four oxidized 20 other organic compounds that fall into those same categories. However, only some of the four Halomonas strains could oxidize the 33 remaining organic compounds assayed (Table 4).

Table 4.  Organic compounds oxidized by the four Halomonas strains isolated without selecting for halotolerance according to Biolog plate substrate utilization matrices
 E-plume1E-plume2E-plume3E-sulfide1H. pacifica
Glycogen   + 
Tween 80+   +
N-Acetyl-D-glucosamine  + +
D-Mannose  +  
D-Psicose   ++
L-Rhamnose  ++ 
Methyl pyruvate   ++
Mono-methyl succinate+  ++
Acetic acid++ ++
cis-Aconitic acid   ++
Formic acid+  ++
D-Galactonic acid lactone  ++ 
D-Galacturonic acid   + 
D-Glucosaminic acid  ++ 
α-Keto glutaric acid  +++
Propionic acid++ ++
Sebacic acid+  + 
Succinamic acid  +++
Glucuronamide   + 
D-Alanine+ +++
L-Alanyl-glycine  +++
L-Aspartic acid  +++
Glycyl-L-glutamic acid  ++ 
Hydroxy L-proline+++ +
L-Leucine   ++
L-Ornithine  ++ 
L-Phenylalanine   ++
L-Pyroglutamic acid +++ 
L-Serine ++++
Inosine   ++
Uridine   ++
Data from H. pacifica shown for comparison.


Halomonas is a cosmopolitan microbial genus, its range encompassing the oceans, sediments, lakes and subsurface environments [33,34]. It not only has penetrated most every environment on Earth, but the results presented here indicate that it is also an abundant group in North and South Pacific pelagic and deep-sea, hydrothermal-vent waters. Less is known about the distribution of Marinobacter, though it is probably ubiquitous given that strains have been cultured from French Mediterranean coast sediments, from deep-sea sediments in the Western Pacific, from oil wells off of the coasts of Vietnam and California, and from the tropical Pacific Ocean off of the Hawaiian coast [33,35–39]. All of the 14 microbial strains cultured from different MPN dilution-series enrichments were phylogenetically identified as Halomonas or Marinobacter, and therefore these halotolerant genera were likely well represented in the majority if not all of the remaining 29 MPNs. Regardless, halotolerant microorganisms, Halomonas, Marinobacter and otherwise, comprised 0.01–10% (order of magnitude) of the microbial community in the Pacific Ocean and its seafloor hydrothermal-vent environments.

The percentages that halotolerant microorganisms constituted of the total microbial community may be inflated due to cell loss while the fixed samples were stored at 2°C for 9–14 months. The maximum estimated cell loss could approach one order of magnitude [40]. However, nearly all counts are in agreement with literature values for marine waters. In the hydrothermal-vent environment, the concentration of microorganisms in lateral hydrothermal plumes and buoyant hydrothermal plumes 2–20 m above a black smoker has been estimated at (11–130)×104 ml−1, and the concentration in bottom water in close proximity to vents was found to be (2–130)×104 ml−1[12,41]. Low-temperature hydrothermal-vent fluids contain (1–11)×104 ml−1[42]. Of all enumerations reported here, only three buoyant hydrothermal plume counts fall below these ranges.

Halomonas and Marinobacter appear to be numerically significant and ubiquitous in marine waters. Typically <0.01–0.1% of the heterotrophic microbial population in a given seawater sample is culturable, and such percentages are achieved by employing numerous types of media incubated under a variety of conditions [5]. It is therefore striking that Halomonas spp. and Marinobacter spp. are so prevalent and so readily cultured, implying that media formulated to mimic the environment may not always be the only approach to culture numerically significant, metabolically active microorganisms.

Our enrichment medium differs from traditional oligotrophic and copiotrophic heterotrophic media in several ways. For the 65 oligotrophs isolated without selecting for halotolerance, there were elevated levels of reduced Fe and Mn, frequently a reduced gas headspace (CH4 or H2/CO2), and trace elements provided in concentrations modeled on 350°C hydrothermal fluid [11], frequently 10–10 000 times higher than measured in seawater. In addition to these factors, abundant halotolerant microbes from MPN enrichments were isolated of course with elevated levels of NaCl, roughly five times higher than seawater. It is unclear which of these factors increased success in culturability, or the mechanism(s) by which these media components operate.

Four Halomonas strains were also obtained when enriching for oligotrophs without selecting for halotolerance from hydrothermal plumes and sulfide rock. While these four strains were identified phylogenetically, additional Halomonas strains may be present among the other 37 oligotrophs obtained without selecting for tolerance to salt. In particular, it is possible that some of the 16 other strains (for a total of 20, 31% of the 65 original isolates) belong to Halomonas because they could also tolerate at least 22% NaCl and had a rod morphology. While high-organic, high-salt media are typically used to select for Halomonas spp. [34], these four strains (and perhaps some of the 16 others) are another example of the infrequent instance when Halomonas spp. are isolated with seawater salinity media. It is the first instance, however, when Halomonas spp. have been isolated with an oligotrophic medium. Again, it is not known whether the elevated levels of Fe, Mn and trace elements or the reduced gas headspace in the enrichment medium enabled Halomonas to out-compete other heterotrophs.

The 16S rRNA phylogenetic tree of Halomonas and Marinobacter environmental isolates reveals clusters of closely related organisms. Mirroring the SAR groups [3,4], strains of nearly identical Halomonas and Marinobacter according to 16S rRNA gene sequence were found thousands of kilometers apart. For example, while nearly identical in 16S rRNA sequence, strains S-sw1, S-sw2 and S-lthf1 were found hundreds of kilometers apart at 21.5°S (2000 m depth), 18.5°S (1000 m depth) and 17.5°S (2570 m depth, low-temperature hydrothermal fluid), respectively, in the Pacific Ocean. Conversely, Halomonas strains with different 16S rRNA sequences were found in close proximity to each other. Strains S-lthf1 and S-lthf2 were isolated from water samples roughly 1 m apart at a low-temperature hydrothermal fluid site at 17.5°S on the SEPR. The geographic distribution of a single cluster is phenomenal. The Marinobacter cluster containing isolates A-sw1, A-sw2, S-plume1, S-plume2 and S-plume3 came from 21.5°S and 46°N, from 10–2600 m depth, and from hydrothermal plume and non-plume environments.

Another important aspect of these data is that cultured organisms do indeed form clusters, thereby allowing the determination of the range of phenotypes within a cluster in conjunction with DNA–DNA homologies. Interestingly, the Biolog plate data indicate that strains E-plume1 and E-plume2 had 92% identity in their ability to metabolize 95 organic compounds, whereas there is only 79% and 74% identity between strain E-plume1 and the more closely related strains E-plume3 and E-sulfide1, respectively. In addition, strain E-plume1 had only 59% identity with H. pacifica, similar to its 58% identity with strain E-twt2, a Vibrio isolate. Using Biolog plates as a metric, there appears to be as much metabolic variation within a cluster as between a cluster and other organisms. This metabolic diversity among closely related Halomonas spp. corroborates the phenotypic variability found in the phylogenetically close-knit cluster of H. aquamarina, H. meridiana and H. variabilis when examining polar lipid patterns, respiratory lipoquinones and proportions of fatty acids [43].

The abundance and ubiquity of Halomonas in marine waters and heavy metal-enriched, low-temperature hydrothermal environments may be attributed to its metabolic and physiological versatility. The four characterized Halomonas strains oxidized 52–74% of 95 organic compounds, grew with 0.002–0.1% organic carbon, grew between −1 and 40°C, utilized oxygen or nitrate, and tolerated 0.05–3 mM Cd2+. These traits are generally consistent with many other characterized Halomonas spp. [34]. The temperature range, ability to grow without oxygen, and metabolic versatility encompass the conditions encountered in the modern ocean; in addition, tolerance to heavy metals permits growth in low-temperature hydrothermal fluids which usually derive from a mixture of hot, metal-laden hydrothermal fluid and seawater. Accordingly, Halomonas should be able to grow under most oceanic and low-temperature, hydrothermal-vent conditions.

Halomonas and Marinobacter have neither been previously cultured from the vent environment nor been seen as numerically important in community phylogenetic studies in the water column [3]. However, most numerically dominant microbes are currently unculturable by traditional methods [7]. While the oligotroph Sphingomonas sp. strain RB2256 was proposed to comprise 15–35% of the pelagic microbial community in an Alaskan bay [44], its putative high abundance was neither confirmed by fluorescence in situ hybridization (perhaps due to low ribosome content) nor by quantitative enrichments [5]. Ultimately, however, Sphingomonas spp. were frequently isolated (∼20% of cellular clones) in a coastal seawater sample [7] though its abundance in open-ocean and deep-sea waters remains unknown. Different regions of the ocean may harbor different populations of numerically dominant microorganisms.

The observation that Halomonas and Marinobacter are abundant members in the water column and in low-temperature hydrothermal fluids raises questions about the role of tolerance to high levels of salt in microorganisms that may experience little variation in salt concentration in the environment. In addition to natural and artificial high-salt environments found along coastlines and in bays (e.g. hypersaline lagoons, solar salterns) and unique settings like the deep Red Sea hot brines [45], there is hypothesized to be an even more pervasive, high-salt environment: brines beneath the global network of mid-ocean ridges.

Brines derive from two processes in deep-sea, mid-ocean ridge hydrothermal systems. The first process, phase separation, occurs at temperatures and pressures above 407°C and 298 bar whereby hydrothermal fluids are transformed into immiscible volatile-rich vapors and droplets of metal-rich brines [46,47]. The phases physically segregate as they circulate within the fracture network of the oceanic crust [48,49]. The Endeavour Segment of the JdFR and 9°N on the East Pacific Rise host robust hydrothermal systems that have been undergoing phase separation as long as observations have been made [50–53]. The second process by which brines may form at depth in submarine environments is by exsolution of magmatic fluids during the final stages of melt crystallization, creating fluids of up to 70% NaCl [47,54–58]. Fluid inclusion analyses of rocks from the Troodos ophiolite and Mid-Atlantic Ridge indicate that hydrothermal fluids with up to 20% NaCl circulate within the lower crustal lithologies in open fracture networks at temperatures of 200–500°C [54,59]. Based on vent fluid compositions and fluid inclusion data, a briny subsurface environment within the gabbroic and lower dike sequences of the oceanic crust was proposed [60]. Here, brines are generated by supercritical phase separation (and possibly degassing magmatic fluid) and subsequently accumulate and persist [61]. Supercritical phase separation and segregation processes are strongly reflected in vent fluid chemistry [50,62,63] in which chlorinities of one-tenth to greater than twice seawater (1–7% NaCl equivalent) are commonly measured in mid-ocean ridge hydrothermal vents [11]. Once cooled or mixed with seawater in a low-temperature hydrothermal fluid system, these brines may create a salty, metal-rich subseafloor microbial habitat. Halomonas, Marinobacter and other halotolerant genera have previously been isolated from coastal oil well and terrestrial subsurface brines [36,37,64,65].

These data are the first report indicating that halotolerant bacteria comprise a significant component of the microbial community in hydrothermal-vent and pelagic marine environments. The ability to grow with high concentrations of NaCl and Cd2+, anaerobically, on a variety of organic compounds, and over a wide mesophilic temperature range implies that the subseafloor brine habitat associated with deep-sea, hydrothermal-vent systems may be a globe-encircling biotope suitable for Halomonas, Marinobacter and other halotolerant and halophilic bacteria and archaea.


This research required the aid of numerous fellow scientists, ships and submersibles. We would like to express our gratitude to chief scientists John Delaney, Bob Embley and Marv Lilley for enabling ample sample collection and to the crews of the R/V Thomas G. Thompson, R/V Atlantis, DSV Alvin and ROV ROPOS. A special thank-you is extended to Debbie Kelley, Dave Butterfield, Jim Holden, Melanie Summit and Byron Crump for their helpful suggestions and advice in the laboratory. This research was supported by Sea Grant (NA36RG0071) and the National Science Foundation (BCS9320070) to J.A.B.