SUPPORTING INFORMATION
Microbial Community Composition and Endolith Colonization at an Arctic
Thermal Spring Are Driven by Calcite Precipitation
Verena Starke, Julie Kirshtein, Marilyn L. Fogel and Andrew Steele
CONTENTS:
Supplementary Text
• Water Chemistry of Troll Springs
• Comparison of Troll to other geothermal springs
Supplementary Materials and Methods
• Sample collection, handling and storage
• PCR amplification for 454 pyrosequencing
• PCR amplification for Sanger sequencing
Supplementary Figures
• Figure S1: Schematic drawings and photographs of sample sites
• Figure S2: SEM images of the Source periphyton
• Figure S3: SEM of entrapment of eukaryotic cells in calcite
• Figure S4: Taxa distribution and abundances of Sanger clone sequences
• Figure S5: SEM of filamentous bacteria in EPS
• Figure S6: Dissolving diatoms in endolithic samples.
• Figure S7: Multi-year pH-temperature correlation data
• Figure S8: OTU distribution at 0.10 distance
Supplementary Tables
• Table S1: Environmental parameters.
• Table S2: db-RDA results
Supplementary References
SI TEXT: ADDITIONAL INFORMATION
Water Chemistry of Troll Springs
The sodium, potassium, sulfate and silicon concentrations in the water at Troll
Springs appear to be controlled by near-equilibrium water-rock interactions,
whereas chloride salinity is possibly derived from fossil sea-water or evaporitic
deposits (Banks et al., 1998). The gases emanating at the source are dominated
by N2 (ca. 70 vol%) and CO2 (25-30%), but also contain considerable amounts of
noble gases (Jamtveit et al., 2006). Nitrate levels are comparatively low in the
source itself and in the most distal pools (Hammer et al., 2005).
Comparison of Troll to other geothermal springs
Active depositing springs can incorporate organic matter (e.g. bacteria,
heterotrophic microbes and eukaryotic algae) into the developing matrix.
Compared to travertine precipitation at Troll Springs, silica can also be deposited
in a geothermal setting. Siliceous sinters are found around hot springs and
geysers, many of which discharge water at or close to boiling. Siliceous sinters
that support microbial communities have been found in New Zealand (Jones et
al., 1997), Iceland (Konhauser et al., 2001; Tobler and Benning, 2011; Tobler et
al., 2008), Tibet (Lau et al., 2008) and Yellowstone (Guidry and Chafetz, 2002;
Inagaki et al., 2001). Like the microbial communities at carbonate springs,
communities tend to be zoned according to temperature (Lau et al., 2008)
(Tobler and Benning, 2011).
Extreme temperatures in geothermal settings can limit the presence of
organisms. For example, Blank et al. (Blank et al., 2002) studied seven silicadepositing springs at Yellowstone with temperatures close to the boiling point.
Streamers in those hot springs are composed of thermophilic organisms with
primary production driven by chemoautotrophic hydrogen oxidation.
Cyanobacteria and green algae are absent, presumably because those
temperatures are above the limit for photosynthetic organisms.
Some silica-precipitating hot springs (e.g. Yellowstone, Iceland, New Zealand)
have an abundant diversity of alkalithermorphilic (pH 7-9 and temperatures 6095ºC) microorganisms, which are mostly chemolithoautotrophic. Key organisms
in those environments include Aquificales, Thermus, Deinococci, Chloroflexus,
Sulfolobus and Synechococcus. Cyanobacteria are typically found only where
the waters have cooled below 70ºC.
Primary production in other Arctic springs has been studied by Perreault (2007;
2008). However, their work concentrated nonphotosynthesis-based primary
production in cold saline sulfide-rich springs, where no cyanobacteria were
found. In contrast, at Troll Springs primary productions seems to be
photosynthetic.
SI MATERIALS AND METHODS
Sample collection, handling and storage
Periphyton samples were collected in sterile Falcon tubes, separated into sterile
Eppendorf tubes, and frozen at -80ºC. Granular and rock samples were collected
in sterile Whirlpacks and stored at -20ºC.
PCR amplification for 454 pyrosequencing
Universal primers 27F and 338R were used for PCR amplification of the V1–V2
hypervariable regions of 16S rRNA genes. The 338R primer included a unique
sequence tag to barcode each sample. The primers were: 27F-5′GCCTTGCCAGCCCGCTCAGTCAGAGTTTGATCCTGGCTCAG-3′ and 338R5′-GCCTCCCTCGCGCCATCAGNNNNNNNNCATGCTGCCTCCCGTAGGAGT3′, where the underlined sequences are the 454 Life Sciences FLX sequencing
primers B and A in 27F and 338R, respectively, and bold letters denote the
universal 16S rRNA primers 27F and 338R. The eight Ns denote the 8-bp
barcode within primer 338R.
The 50 µl reaction mixture contained 10 mM (total) deoxynucleoside
triphosphates (dNTPs), 0.5 µM (each) primer, 5 ng/µl of DNA template, 0.5 U of
Phusion High-Fidelity DNA polymerase (New England BioLabs), and 5 x Phusion
PCR buffer HF, containing 7.5 mM MgCl2, and 3% DMSO. PCR conditions were
1 cycle of 30 seconds at 98°C, followed by 30 cycles of 5 seconds at 98°C, 15
seconds at 55°C, and 45 seconds at 72°C using a DNA Engine DYAD PCR
machine (MJ Research). 10-min incubation at 72°C was the final step. Negative
controls without a template were included for each barcoded primer pair.
Sequencing was done at the Genomics Resource Center at the Institute for
Genome Sciences (IGS), University of Maryland School of Medicine, using
protocols recommended by the sequencing system manufacturer as amended by
the Center. The concentrations of amplicons were estimated using a GelDoc
quantification system (Bio-Rad Laboratories), and approximately equal amounts
(100 ng) of all amplicons were mixed in a single tube. Amplification primers and
reaction buffer were removed using the AMPure Kit (Agencourt, Beckman
Coulter Genomics). Emulsion PCRs were performed as described in Margulies et
al. (Margulies et al., 2005). Sequences were obtained using a Roche 454 GSFLX sequencing system (Roche-454 Life Sciences). Assessment of the quality of
the sequences and binning using the sample-specific barcode sequence was
performed at IGS.
PCR amplification for Sanger sequencing
16S rRNA genes of 10 samples (Fig. S4) were amplified by PCR using the
B27/1492R primer set. The amplified DNA fragments were gel-purified, cloned
and sequenced in both directions (M13F/M13R primers) by Macrogen Inc (Seoul,
Korea) using an ABI3730 XL DNA Analyser (Applied Biosystems, Renton, USA).
Ninety-six clones from each clone library were randomly picked for sequence
analysis.
Any use of trade, firm, or product names is for descriptive purposes only and
does not imply endorsement by the U.S. Government.
SI FIGURES
Figure S1: Schematic drawings and photographs of sample sites.
Photos: V. Starke/K.O. Storvik/AMASE
Figure S2: SEM images of filamentous algae in the source periphyton at low (a)
and high (b) resolution.
Figure S3: Entrapment of eukaryotic cells in calcite. Calcite crystals (a, b)
ultimately merge to form a contiguous sheath (c) that accumulates around
eukaryotic algae cells. White arrows point to calcite crystals and sheath.
Figure S4: Taxa distribution and abundances of Sanger clone sequences for 10
Troll samples. The grid displays the proportional distribution of each taxon across
all samples (not the proportion in each sample). The abundances are shown in
the bar graph.
Figure S5: SEM of filamentous bacteria (arrows) in EPS. Both images are etched
granular samples showing that filamentous bacteria are prominent but
filamentous eukaryotic algae are less common.
Figure S6: Dissolving diatoms in endolithic samples. Diatom frustule dissolution
is well documented in a variety of settings (Ryves et al., 2001; Flower, 1993;
Loucaides et al., 2008). Dissolution rate is dependent on pH, salt concentration
and temperature of surrounding media (Ryves et al., 2001; Flower, 1993). For
example, dissolution rates double as pH increases from 6.3 to 8.1 (Loucaides et
al., 2008), and decrease with decreasing temperature (Tréguer et al., 1989).
Figure S6 shows partially dissolved diatom frustules: an impression in the calcite
(a) and a partially dissolved frustule leaving just a ribcage-like structure behind
(b).
Figure S7: Temperature and pH correlation for data collected over three years at
Troll Springs. The sensitivity to weather varies from pool to pool, and is greatest
for pool 1 (which is broad and shallow) and least for the source (which is narrow
and deep). Warmer pools, such as pool 1, tend to cool more sharply in response
to cold air and winds because of their warmer initial temperatures. Pool 2, in
contrast, has a cooler initial temperature, resulting in smaller temperature and pH
variations.
Figure S8: OTU distribution at 0.10 distance. OTUs are displayed as presence or absence. Only OTUs that are present in
at least two samples are shown.
a: OTUs have been ordered according to the number of shared OTUs in four endolithic samples (starting with four shared
among the four samples, then three among the four, and so on). OTUs shared with other samples, such as periphyton
and granular, are on the left side. All OTUs not shared with the endolithic samples, but shared between periphyton and
granular samples, are on the right side. The terrestrial samples (rows) are ordered according to their water content. Some
OTUs are specific to certain types of samples, but others are present through all samples, indicating that they persist
through the transition from aquatic to terrestrial. This panel also displays a transition (blocks of shared OTUs) from
periphyton to granular and from granular to endolith (red arrow).
b: OTUs have been ordered according to abundances in the Terrace 1 granular center sample. All OTUs shared with that
sample are located on the left. The arrow indicates the transition from the periphyton through the granular into the
endolithic samples. This panel emphasizes OTUs shared with the Terrace 1 center granular sample, which has the
highest diversity and water content of all the terrestrial samples. Because it was recently wet, this sample also represents
a transition between pool and endolithic environments. In general, samples plotted next to Terrace 1 center granular
share more OTUs than the samples farther away, which are also more different in their environmental parameters (e.g.,
source periphyton and driest endolithic samples).
SI TABLE
Table S1: Environmental parameters measured during sample collection. Measurements include water and rock samples.
n.a. = not available. Letters in column 2 are the order of samples and refer to the sample labeling in figure 5, connecting
sample points to corresponding environmental variables.
Samples
Temperature
[°C]
Water Content
[%wt]
pH
Eh [mV]
Dissolved
Oxygen [%]
Conductivity
[uS/cm]
Source mud
A
24.8
58.8
6.65
10
10
1616
Source periphyton 1
B
24.8
69.1
6.65
10
10
1616
Source periphyton 2
C
24.8
66.1
6.65
10
10
1616
Pool 1 periphyton
D
16.7
64.8
7.33
-30
126
1618
Pool 2 periphyton
E
9.4
61.2
8.06
-70
99
1545
Pool 3 periphyton
F
12.7
59.8
7.94
-60
105
1518
Terrace1 center granular
G
6
24.9
n.a.
n.a.
n.a.
n.a.
Terrace 1 edge granular
H
6
10.7
n.a.
n.a.
n.a.
n.a.
Terrace 1 rim granular
I
6
3.3
n.a.
n.a.
n.a.
n.a.
Terrace 2 center endolith
J
4
7.0
n.a.
n.a.
n.a.
n.a.
Terrace 2 rim endolith
K
4
1.9
n.a.
n.a.
n.a.
n.a.
Terrace 3 rim endolith
L
4
1.0
n.a.
n.a.
n.a.
n.a.
Terrace 4 rim endolith
M
4
0.3
n.a.
n.a.
n.a.
n.a.
Table S2: Relationship of microbial community structure to environmental parameters as determined by UniFrac db-RDA.
All principal coordinates were used.
Environmental
variable
One constraint variable, separate models
Terrestrial samples*
Water content
Aquatic samples**
Temperature
pH
Samples
Several constraint variables applied
sequentially, one model***
**
Aquatic samples
Temperature
pH
eH
Proportion of
p-value
variance explained
Total Sequence Set
0.31
0.012
0.58
0.027
0.57
0.016
0.58
0.20
0.20
0.017
0.046
0.082
Proportion of
p-value
variance explained
Bacterial Sequence Set
0.32
0.011
0.38
0.074
0.38
0.066
0.38
0.20
0.19
0.166
0.613
0.693
* terrestrial samples analyzed: granular (3) and endolithic (4) samples
** aquatic samples analyzed: source periphyton (2) and pool periphyton (3) samples. The Source mud sample was excluded.
*** only first three selected parameters shown
SI REFERENCES
Banks,D. et al. (1998) The thermal springs of Bockfjord, Svalbard: occurrence
and major ion hydrochemistry. Geothermics 27: 445–467.
Blank,C.E. et al. (2002) Microbial composition of near-boiling silica-depositing
thermal springs throughout Yellowstone National Park. Applied and
Environmental Microbiology 68: 5123.
Flower,R.J. (1993) Diatom preservation: experiments and observations on
dissolution and breakage in modern and fossil material. Hydrobiologia 269270: 473–484.
Guidry,S.A. and Chafetz,H.S. (2002) Factors governing subaqueous siliceous
sinter precipitation in hot springs: examples from Yellowstone National Park,
USA. Sedimentology 49: 1253–1267.
Hammer,Ø. et al. (2005) Evolution of fluid chemistry during travertine formation in
the Troll thermal springs, Svalbard, Norway. Geofluids 5: 140–150.
Inagaki,F. et al. (2001) Silicified microbial community at Steep Cone hot spring,
Yellowstone National Park. Microbes and Environments 16: 125–130.
Jamtveit,B. et al. (2006) Travertines from the Troll thermal springs, Svalbard.
Norwegian Journal of Geology 86: 387.
Jones,B. et al. (1997) Biogenicity of silica precipitation around geysers and hotspring vents, North Island, New Zealand. Journal of Sedimentary Research
67: 88–104.
Konhauser,K.O. et al. (2001) Microbial–silica interactions in Icelandic hot spring
sinter: possible analogues for some Precambrian siliceous stromatolites.
Sedimentology 48: 415–433.
Lau,C.Y. et al. (2008) Early colonization of thermal niches in a silica-depositing
hot spring in central Tibet. Geobiology 6: 136–146.
Loucaides,S. et al. (2008) Dissolution of biogenic silica from land to ocean: role
of salinity and pH. Limnology and Oceanography 53: 164–1621.
Margulies,M. et al. (2005) Genome sequencing in microfabricated high-density
picolitre reactors. Nature 437: 376–380.
Perreault,N.N. et al. (2007) Characterization of the prokaryotic diversity in cold
saline perennial springs of the Canadian high Arctic. Applied and
Environmental Microbiology 73: 1532–1543.
Perreault,N.N. et al. (2008) Heterotrophic and autotrophic microbial populations
in cold perennial springs of the high arctic. Appl Environ Microbiol 74: 6898–
6907.
Ryves,D.B. et al. (2001) Experimental diatom dissolution and the quantification of
microfossil preservation in sediments. Palaeogeography, Palaeoclimatology,
Palaeoecology 172: 99–113.
Tobler,D.J. and Benning,L.G. (2011) Bacterial diversity in five Icelandic
geothermal waters: temperature and sinter growth rate effects. Extremophiles
15: 473–485.
Tobler,D.J. et al. (2008) In-situ grown silica sinters in Icelandic geothermal areas.
Geobiology 6: 481–502.
Tréguer,P. et al. (1989) Kinetics of dissolution of Antarctic diatom frustules and
the biogeochemical cycle of silicon in the Southern Ocean. Polar Biology 9:
397–403.