Geobiology (2011), 9, 411–424
DOI: 10.1111/j.1472-4669.2011.00288.x
Hot spring siliceous stromatolites from Yellowstone National
Park: assessing growth rate and laminae formation
W. M. BERELSON,1 F. A. CORSETTI,1 C. PEPE-RANNEY,2 D. E. HAMMOND,1 W. BEAUMONT1
AND J. R. SPEAR2
1
Department of Earth Sciences, University of Southern California, Los Angeles, CA, USA
Division of Environmental Science and Engineering, Colorado School of Mines, Golden, CO, USA
2
ABSTRACT
Stromatolites are commonly interpreted as evidence of ancient microbial life, yet stromatolite morphogenesis is
poorly understood. We apply radiometric tracer and dating techniques, molecular analyses and growth experiments to investigate siliceous stromatolite morphogenesis in Obsidian Pool Prime (OPP), a hot spring in Yellowstone National Park. We examine rates of stromatolite growth and the environmental and ⁄ or biologic conditions
that affect lamination formation and preservation, both difficult features to constrain in ancient examples. The
‘‘main body’’ of the stromatolite is composed of finely laminated, porous, light–dark couplets of erect (surface
normal) and reclining (surface parallel) silicified filamentous bacteria, interrupted by a less-distinct, well-cemented ‘‘drape’’ lamination. Results from dating studies indicate a growth rate of 1–5 cm year)1; however, growth
is punctuated. 14C as a tracer demonstrates that stromatolite cyanobacterial communities fix CO2 derived from
two sources, vent water (radiocarbon dead) and the atmosphere (modern 14C). The drape facies contained a
greater proportion of atmospheric CO2 and more robust silica cementation (vs. the main body facies), which we
interpret as formation when spring level was lower. Systematic changes in lamination style are likely related to
environmental forcing and larger scale features (tectonic, climatic). Although the OPP stromatolites are composed of silica and most ancient forms are carbonate, their fine lamination texture requires early lithification.
Without early lithification, whether silica or carbonate, it is unlikely that a finely laminated structure representing
an ancient microbial mat would be preserved. In OPP, lithification on the nearly diurnal time scale is likely related
to temperature control on silica solubility.
Received 14 January 2011; accepted 15 June 2011
Corresponding author: W. M. Berelson. Tel.: 213 740 5828; fax: 213 740 8801; e-mail: berelson@usc.edu
INTRODUCTION
Stromatolites are most commonly defined as laminated organo-sedimentary structures built by the trapping and binding
and ⁄ or precipitation of minerals by microbial mats (e.g., Walter, 1976; Riding, 1990; Riding & Awramik, 2000; Awramik
& Grey, 2005). As such, their presence is commonly taken as
a biosignature, and they are presumed to constitute some of
the oldest evidence for life on Earth (Hofmann et al., 1999;
Allwood et al., 2006). However, morphology can be deceiving. Abiotic structures may mimic biogenic stromatolites
(e.g., Lowe, 1994; Mcloughlin et al., 2008), and numeric
modeling demonstrates that stromatolite morphogenesis can
be modeled with simple rules – life may or may not be
required to build the domical, columnar, and ⁄ or branching
forms typical of most Precambrian stromatolites (e.g., Grotzinger & Rothman, 1996; Grotzinger & Knoll, 1999; Dupraz
2011 Blackwell Publishing Ltd
et al., 2006). The biogenicity of ancient stromatolites is not
assured and the forcing responsible for lamination is largely
unknown.
Few modern stromatolites have been age dated in order to
understand their growth rate, and for the age-dated specimens, accretion rates do not match daily, seasonal or yearly
lamination accretion most commonly cited for ancient stromatolite growth (e.g., Chivas et al., 1990, Petryshyn et al.
2011). The paucity of actively growing modern examples,
whether biogenic or abiogenic, make meaningful studies with
respect to growth rate, morphology, biotic control (if
present), environmental forcing, etc., difficult. Well-studied
modern marine examples are known and include Shark Bay,
Australia, (e.g., Logan, 1961; Hoffman, 1976) and the Bahamas (e.g., Dravis, 1983; Dill et al., 1986; Reid et al., 2000).
However, most modern marine forms are crudely laminated,
if laminated at all, and while they are instructive for our
411
412
W. M. BERE LS ON et a l.
understanding of microbialite growth in general, do not constitute a good textural (lamination scale) analog for the typically finely laminated Precambrian stromatolites (cf., Awramik
& Riding, 1988; Grotzinger & Knoll, 1999; Awramik &
Grey, 2005).
Here, we present data from a living stromatolite from
Yellowstone National Park that provides a textural analog to
many Precambrian forms and describe experiments conducted
to constrain stromatolite growth rate. Furthermore, we
discovered 14C is useful as a tracer for CO2 uptake in this
system, which, as we will demonstrate, has led to a better
understanding of stromatolite morphogenesis in this system,
with relevance to the rock record. We also discuss results of hot
spring temperature fluctuations and other environmental forcing as relevant to the mineralization of Yellowstone stromatolites.
BACKGROUND
Yellowstone National Park (YNP) continues to be a locus of
geomicrobiological study. The spectrum of environments and
microbiota within and around hot springs is astounding, with
temperatures ranging from ambient to 94 C (surface boiling
at 2280 m; temperatures higher at depth in springs), and
pH values between 0 and 10; each unique set of conditions
provides a niche for a unique community of micro-organisms.
Silica-rich hot springs provide an excellent environment for
stromatolite morphogenesis for several reasons (as noted in
Walter et al., 1972). First, ‘‘harsh’’ acidic and ⁄ or high temperature conditions exclude most metazoan life that might
graze on the microbial communities and disrupt stromatolite
formation. Second, and perhaps most importantly, dissolved
silica will adsorb and precipitate on microbial mats as the
spring water cools – regardless of the microbial metabolism –
providing a situation where very early lithification of microbial
mats can occur. And thirdly, silica solubility is a strong
function of temperature and hot spring waters are regularly
subjected to temperature fluctuations due to seasonal weather
conditions and flow in shallow channels. There have been
numerous studies of YNP microbial mats, sensu stricto (e.g.,
Ferris et al., 2003; Allewalt et al., 2006; Steunou et al., 2006,
2008; Bhaya et al., 2007; Bryant et al., 2007; Klatt et al.,
2007; van der Meer et al., 2007; Schouten et al., 2007).
However, a typical microbial mat is not likely to end up in the
rock record as a stromatolite. The mat must be lithified or
mineralized in some way to enter the fossil record as a stromatolite, and silica-rich hot springs provide an excellent environment for early lithification to occur. Although many
ancient stromatolites are composed of CaCO3, the process of
microbial growth plus early lithification (in this case, via
silicification) forms an excellent textural analog for ancient
stromatolite morphogenesis.
Stromatolites have been recognized in YNP for decades
(e.g., Barghoorn & Tyler, 1965; Walter et al., 1972;
Doemel & Brock, 1974; Awramik & Vanyo, 1986; Guidry &
Chafetz, 2003). Additionally, several recent reviews discuss
microbial hot spring lithification and its effect on sinter laminations and microstructure (Konhauser et al., 2003, 2004;
Benning et al., 2005). Walter et al. (1972) give credit to
Weed (1889) for first describing stromatolites in YNP
(although the biogenic nature was not known at that time).
Barghoorn & Tyler (1965) recognized that the microstructure of YNP stromatolites was similar to some Precambrian
forms (in particular, the ca. 1.88 Ga Gunflint Formation
stromatolites, which are composed of silica and contain
microbial fossils). Laminated microbialite morphogenesis in
YNP occurs in many ways, from subaqueous lithification of
microbial mats (e.g., Walter et al., 1972; Walter, 1976) to
subaerial, intermittently splashed microbial mats (e.g., Cady
& Farmer, 1996; Blank et al., 2002). Hinman & Lindstrom
(1996) describe the lithification process as related to wicking
and evaporation but also to changes in solubility as forced by
changes in temperature. The stromatolites we describe here
are microstructurally similar to the structures described by
Walter et al. (1972) and Walter (1976) in that the laminations are composed of silica-coated filamentous bacterial
sheaths, where the filaments are erect in one layer and reclining in the next. Our samples differ in that they are domal or
columnar on the mesoscopic scale (sensu Shapiro, 2000)
rather than conical. The stromatolites we study here have
two lamination structures, a ‘‘main body’’ of high porosity,
fine laminae couplets and a ‘‘drape’’ structure that has low
porosity and is denser, as described below.
SITE DESCRIPTION AND METHODS
Obsidian pool prime
Samples were collected from a small hot spring located to the
west of the Mud Volcano area (within Hayden Valley) known
as Obsidian Pool Prime (OPP, Fig. 1A), located adjacent to
Obsidian Pool hot spring (Spear et al., 2005; Shock et al.,
2005; Hugenholtz et al. 1998). The spring itself is 600 m2
and has a pH of 5.7. It receives a small fraction (<10%) of
inflow from Obsidian Pool, primarily filling and overflowing
with water from its own source vent. The temperature of
the incoming vent water is 75 C, and it cools to 45 C
(measurements made in summer months) as it travels from
the vent to the northeastern rim of the pool where the
drainage outflow is located. The spring first received attention
because of its H2 concentration – among the highest in YNP –
and the microbial communities associated with H2 metabolisms (e.g., Spear et al., 2005). Venting occurs in the center of
the spring and toward the southern rim. The dissolved Si
content of OPP spring water averages 5.7 mM (for comparison, the adjacent Obsidian Pool averages 2 mM). Gas bubbles
from the water from Obsidian Pool are 98% CO2 (Shock
et al., 2005), likely originating from the dissolution of subsurface Paleozoic limestones or perhaps directly from volcanic
2011 Blackwell Publishing Ltd
Siliceous stromatolite growth rates
413
A
Shelf
RS
B
C
S
RS
Fig. 1 Field photographs of Obsidian Pool Prime (OPP) and associated stromatolites. (A) Stromatolites grow as isolated bodies along the shelf or as a semi-continuous
rim on the north side of the pool (RS, rim stromatolite). (B) Close-up of inset box in A, with rim stromatolites (RS) and isolated stromatolites (S). Note depression in the
center of the isolated stromatolite (a common feature of these stromatolites). (C) Recently harvested isolated stromatolite.
sources. The stromatolites are concentrated within OPP along
the northern and western rim of the spring (Fig. 1).
Samples
Three small stromatolites were removed from OPP for the
purposes of conducting the analyses described here (microscopy, elemental composition, radioisotope analysis, and
molecular biological analysis). Stromatolite (I) was collected
in May 2006 from within the spring and stands approximately
6 cm tall and 10 cm in diameter in a mushroom-shape (Figs 1
and 2). It was submerged 15 cm below the water–atmosphere
interface and located approximately 1 m into the spring from
the spring-land border. Stromatolite (II) was collected in
October 2006 from a location at the spring–land boundary.
This structure apparently originated on a cobble, was also
approximately 6 cm tall and 12 cm in diameter and was similarly structured to Stromatolite (I) with a laminated main body
and two intervals where the main body growth is interrupted
by a ‘‘drape’’ structure. Stromatolite (III) was ‘‘grown’’ on an
inverted Erlenmyer flask, emplaced within OPP on October 3,
2008 and collected on February 9, 2009. In all cases, the stromatolites were air dried and cut with a scalpel or razor to reveal
fine structure and provide clean surfaces for sampling.
2011 Blackwell Publishing Ltd
SEM
Scanning electron microscopy was performed on samples
obtained from Stromatolite (I) using a Hitachi (Tokyo, Japan)
TM-1000 benchtop eSEM with an Oxford Elemental Analysis
(EDS) unit.
14
C-Radiocarbon analyses were performed at the Keck
Carbon Cycle AMS Facility at the University of California,
Irvine. As is convention, radiocarbon concentrations are given
as D14C and ages calculated following Stuiver & Polach
(1977). Samples from Stromatolite (II) were acid washed to
ensure that no adsorbed carbon was analyzed and sample
preparation backgrounds have been subtracted, based on
measurements of 14C-free coal. All results have been corrected
for isotopic fractionation according to the conventions of
Stuiver & Polach (1977).
Th, Ra, Cs isotopes
Samples from Stromatolites (II and III) were analyzed for
thorium-228 (based on daughter radium-224), radium228 (based on daughter actinium-228), radium-226
(based on daughters lead-214 and bismuth-214), and
cesium-137 isotopes on a gamma spectrometer using a
414
W. M. BERE LS ON et a l.
Molecular analyses
Living mat
A
F
E
G
H
Main
body (2)
D
C
Drape
B
Main body (1)
Fig. 2 Siliceous stromatolite obtained from the OPP hot spring, YNP. Scale
units at bottom of frame are 1 mm. Letters indicate sites where C, N, and S
measurements were made (see Table S1). Note two distinct styles of lamination: ‘‘main body’’ and ‘‘drape.’’ Following the growth of the main body (1),
growth was replaced by the drape fabric; growth returned to the main body
style of lamination (main body 2). When collected, a living mat – an incipient
drape – topped the structure.
low background counting system (EGG Ortec) with a
well-type intrinsic germanium detector. For these analyses,
1 g of dry material was powdered and placed in a test
tube inserted into the detector. Uncertainties reported are
based on counting statistics.
Samples for molecular analyses were collected by scraping
mats from stromatolite surfaces with sterile razor blades. Samples of ‘‘main body’’ material was obtained from Stromatolite
(III), ‘‘drape’’ material was obtained from Stromatolite (I).
Samples were transferred to cyrovials and frozen in liquid
nitrogen within 1 h of sampling for transport to the lab. Samples were kept at )80 C for long-term storage. DNA was isolated from each sample using the MoBio PowerSoil DNA
Extraction Kit (Carlsbad, CA, USA). DNA was amplified
using bacterial primers 8F (Lane, 1991) and 338R (Amann
et al., 1995). Eight base barcodes to denote each sequence’s
origin and sequencing adapters were incorporated into each
primer as described previously (Hamady et al., 2008).
Sequencing was done on the Roche 454 (Branford, CT,
USA) GSFLX pyrosequencing platform. Sequences were
de-noised using Denoiser version 0.851 (Reeder & Knight,
2010) and binned by barcodes using the QIIME software
package (Caporaso et al., 2010). Sequences were clustered
using the UClust (Edgar, 2010) wrapper in QIIME. Figures
depicting sequence distributions were created using custom
Python scripts that incorporated modules in Matplotlib (Hunter, 2007). Phylum level classifications of pyrosequences were
found by recruiting reads to taxonomically annotated reference
sequences in the Siva SSURef104 SSU rRNA gene database
using the NCBI BLAST (Altschul et al., 1990) wrapper in
QIIME. Near neighbors to relevant pyrosequences were found
by BLAST searching against the Silva SSURef104 database.
Growth experiments
In June and early October 2008, we placed an inverted
Erlenmyer flask in OPP within arm’s reach of the shoreline. The flat bottom of the flask was roughly equivalent
to the surface level of the pool or a few centimeters submerged. The flask was held in this position with a stake.
The June flask was recovered in September and the October flask recovered in February 2009. The second growth
experiment had significant growth, which in all respects
resembled Stromatolites (I) and (II), and this structure
was deemed Stromatolite (III).
Temperature fluctuations
Variability in pool temperature was monitored hourly
between June 2010 and February 2011 using ibutton
(Maxim, Sunnyvale, CA, USA) temperature sensors.
These devices are temperature recorders, cylindrical in
shape, 2 cm diameter and 4 cm long. They were placed
within water-resistant cases and mounted on stakes within
arms-length of the shoreline and placed at the N, S,
and W portions of OPP submerged to 10 cm depth.
We report the temperature fluctuations from the Northern location; however, all temperature records were
identical.
RESULTS AND DISCUSSION
Stromatolite morphology
Stromatolites occur as a continuous rim along portions of
OPP or as isolated mushroom-shaped structures on the shallow shelf adjacent to the rim (Fig. 1). The isolated specimens
range from a few centimeters to 30 cm across their top and
can be 10–20 cm tall (Fig. 1C). A transverse section through
Stromatolite (I) reveals the distinct laminated fabric and complex internal pattern of convex upward accretion (Fig. 2). This
morphology and structure is common to all the stromatolites
examined from OPP. A cross-section through the rim stromatolites reveals upward growth first (similar to the isolated
stromatolites), followed by lateral accretion toward the center
of the pool. Two distinct styles of lamination are observed in
the stromatolites (Fig. 2) – the most common style comprises
the majority of the stromatolite (the ‘‘main body,’’ Fig. 2),
while the other style is less common and less distinctly laminated (the ‘‘drape,’’ Figs 2–4).
The majority of the main body of the stromatolite is composed of alternating light and dark laminations that are themselves comprised of simple, unbranching silica-coated
filamentous tubes (Fig. 3). Laminae range in thickness
2011 Blackwell Publishing Ltd
Siliceous stromatolite growth rates
415
A
Dark
Light
100 µm
C
B
Light
Light
Dark
60 µm
Dark
40 µm
Fig. 3 Relevant features of the main body of the stromatolite. (A) Banding of stromatolite seen as alternating layers of silicified tubes, one set oriented primarily in the
direction of growth (toward top of page), forming a light-colored lamination in Fig. 2, and the next layer oriented perpendicular to growth, forming a dark lamination.
(B) Closeup on the left flank of a stromatolite. (C) Some light colored layers reveal large ‘‘cavity’’ structures, which we interpret as the former presence of gas bubbles
in the mat. Similar structures are visible in Fig. 3A.
between 100 and 200 lm, but in general, the light layers average 150 lm and the dark layers approximately 100 lm. The
main body remains quite permeable – there is little silica
cement between the filaments in either the light or dark layers.
EDS, X-ray Diffraction and sodium bicarbonate leach analyses
of the laminae indicate that they are >95% amorphous SiO2
(opal). The dark layers are composed of densely packed tubes
oriented sub-parallel to the lamination, and the light layers are
composed of sparsely packed tubes oriented sub-normal to
the lamination. The lighter layers have greater porosity and
many have large cavities (Fig. 3C), suggestive of gas bubbles
within the microbial mat (cf., Bosak et al., 2009).
The main body lamination style is interrupted by a distinctly
different, less well defined lamination style, termed here ‘‘the
drape.’’ It is apparent that the main body of the stromatolite
grew upwards, and at certain intervals, the whole structure –
the sides as well as the top – was draped by a denser, less-
2011 Blackwell Publishing Ltd
distinct form of lamination (Fig. 4). Ultimately, construction
returned to the main body style of lamination and the drape
morphology represents a depositional layer or facies within
the stromatolite (note the two generations of main body lamination in Fig. 2). The drape contains silicified tubes, some
clearly branching, as well as coccoidal forms and pennate
diatoms that are not found in the main body (Fig. 4). Unlike
the light–dark couplets in the main body, there is less internal
structure to the drape, and the tubes are more tightly packed.
The drape pore space is entirely filled with silica (Fig. 4A,D),
in contrast to the weakly cemented main body facies. The
drape and the main body fabrics alternate over time; the outer
portion of the stromatolite in Fig. 2 represents a second
drape, indicating at least two episodes of main body formation
punctuated by two episodes of drape formation. In the
stromatolites studied here, the drape facies is volumetrically
subsidiary to the main body facies.
416
W. M. BERE LS ON et a l.
A
B
20 µm
30 µm
C
D
Drape
Main body
40 µm
400 µm
Fig. 4 Relevant features of the drape fabric. (A) The porosity is commonly occluded by silica in the drape facies, in contrast to the main body, which remains a more
open framework. While the centers of the tubes remain open, the spaces between them are occluded. (B) Diatoms are a common constituent of the drape. (C) Coccoidal cells are also a common constituent of the drape, differentiating it from the main body fabric. (D) A well-cemented drape is clearly visible, vs. the more open structure of the main body in this SEM image.
Stromatolite molecular analysis
We identify the main body tubes as sheaths of a filamentous, non-heterocystous cyanobacterium based on observation of their morphology and molecular analyses. Although
the main body mat contains a non-branching filamentous
cyanobacterial form, the drape facies possesses true-branching, heterocystous morphologies as well as short chains of
coccoidal cyanobacterial cells. Pyrosequence libraries of bacterial 16S amplicons from each mat type show that cyanobacteria are the most abundant phylum in each library (70
and 41% from the main body and drape mat, respectively,
Fig. 5A) although 16 other phyla are also represented. The
cyanobacteria, due to larger cell volume, appear to constitute an even larger proportion of the community biomass
than fraction of pyrosequences. The cyanobacterial members of the main body mat are dominated by one cyanobacterial phylotype while the drape accreting mat
cyanobacterial membership is dominated by one major and
one minor phylotype. The dominant cyanobacterial member of the main body facies does not have significant
(>92%) sequence identity with any other sequence in the
Silva SSURef104 SSU rRNA gene database that is classified
beyond the phylum level (Fig. 5B). The major drape cyanobacterial phylotype is closely related (>99% sequence identity) to the Chlorogloeopsis cultivar Chlorogloeopsis sp. PCC
7518 (Accession X68780, Wilmotte et al., 1993) and the
minor phylotype shows high identity (>99%) with a cultivated Fischeralla species (Accession DQ786171, Finsinger
et al., 2008) from Costa Rica hot springs. Details of the
molecular composition of OPP stromatolites may be found
in Pepe-Ranney et al. (2011).
Stromatolite growth rate experiment
To determine stromatolite growth rate, a substrate was
provided in OPP and growth was monitored with time.
Interestingly, little stromatolite growth was noted on substrates during the summer months. The flask that was emplaced in June had no appreciable growth on it in
October. However, robust growth was noted during the
fall and winter. Stromatolite (III) was grown on the flat
base of an inverted flask over 141 days between October
3, 2008 and February 9, 2009 when the flask was recovered. The structural composition of the body of Stromatolite (III) is identical to the body of Stromatolites (I) and
(II) (Fig. 6). A small portion of the base of Stromatolite
(III) was lost during collection, but the remaining piece,
in section, reveals a sequence of approximately 80 (±10)
couplets of light and dark laminae (taking into account an
approximation of the amount of material lost during sampling). The top of this structure was covered with the
drape fabric.
The overall growth of Stromatolite (III) was approximately
2 cm in half a year. Such rapid growth may be the fortuitous
result of where and how we planted a growth platform, thus
other age dating techniques were applied (below). Moreover,
growth at such a high rate is clearly not a steady-state process,
2011 Blackwell Publishing Ltd
Siliceous stromatolite growth rates
A 1.0
Aquificae
Fibrobacteres
Caldiserica
Gemmatimonadetes
Can. division OP2
Actinobacteria
Planctomycetes
Spirochaetes
Can. division OP11
Acidobacteria
Can. division OP10
Bacteroidetes
Chlorobi
Chloroflexi
Proteobacteria
Deinococcus-Thermus
Cyanobacteria
0.6
B
1 cm
Fig. 6 Image of stromatolite (III) demonstrating the alternate fine-scale lamination within the main body representing couplets of light and dark laminae.
There are approximately 80 ± 10 couplets in this stromatolite which grew during a period of 141 days between October 2008 and February 2009.
100
Radiometric analysis of stromatolite growth rate
(228Th ⁄ 228Ra, 228Ra ⁄ 226Ra, and 137Cs)
0.4
60
0.2
40
20
0
Main body mat
Drape mat
0
sequences
80
% Cyanobacterial
Fraction of sequences
0.8
417
1
2
Rank
3
Main body mat
Drape mat
Fig. 5 (A) Stacked bar chart depicting the phylum distribution of Bacterial 16S
pyrosequence libraries. (B) Rank abundance plot of Operational Taxonomic
Units (OTUs) of cyanobacterial sequences in each library. All sequences in each
OTU have at least 97% sequence identity to the seed sequence for the OTU.
The main body mat library is dominated by one OTU or phylotype while the
drape mat is dominated by two phylotypes.
as indicated by the absence of growth on the June–September
flask experiment and the interruption of main body fabric with
drape fabric. Nonetheless, this simple experiment demonstrated that the OPP stromatolites can form very quickly and
the lamination frequency maybe nearly diurnal during intervals of accretion.
The activity of several radioactive elements and element ratios
were determined to see if different age models were internally
consistent and if this approach might help confirm the stromatolite growth rate established by the flask experiment
described above. Stromatolite (II) was sampled for radiometric analysis from the upper drape surface (Table 1, Sample A),
the middle of the main body (B), and lower portion of the
main body (C). An integrated sample from Stromatolite (III),
the farmed stromatolite, was also analyzed.
We measured two isotope ratios, 228Th ⁄ 228Ra and
228
Ra ⁄ 226Ra, and made the assumption that the stromatolites
contain no 228Th from any source other than 228Ra decay, and
that the 228Ra ⁄ 226Ra isotope ratio is constant in the spring
water. Stromatolite (III) provides a quasi-control insofar as we
know its growth history and can decay-correct its measured
isotope values to the time of collection. Our results on Stromatolite (II) suggest that the upper surface was 1.1–1.5 years
old and that the underlying layers were between 2–6 and
3–8 years old (Table 1). Cesium-137, an isotope produced
during bomb testing in the early 1960s, decreases from
bottom to top, and is thus consistent with our interpretation
that the bottom portion of this stromatolite was formed
post-1960. The consistency between the thorium ⁄ radium,
radium ⁄ radium, and 137Cs dating techniques support the
interpretation that the age of these structures is in years rather
than in hundreds or thousands of years. Stromatolite (III),
grown on a flask over a known time interval, had a
228
Ra ⁄ 226Ra ratio 1.66. This result is consistent with the
Table 1 Radiometric analyses on silicious Stromatolites (II and III)
ID
137
228
±
Age
Range
228
±
Age
Range
II-A
II-B
II-C
III
0.9
2.2
2.6
0.41
0.96
1.52
0.04
0.22
0.44
1.3
4.3
>5
1.1–1.5
2.8–6.5
5.3 to >5
1.40
1.00
0.90
1.66
0.14
0.25
0.27
0.97
1
4
5
2–6
3–8
Cs
Th ⁄ 228 Ra
Ra ⁄ 226 Ra
Cs activity is in bq kg)1. Ages are in years. Range takes into account ±1 SD in counting statistics. 228Ra ⁄ 226Ra age of stromatolite (II) assumes initial ratio (1.40)
represents a 1-year-old sample. A, B, and C refers to the sample location within Stromatolite (II), where A is the shallowest sample and C is the deepest.
2011 Blackwell Publishing Ltd
418
W. M. BERE LS ON et a l.
Table 2
14
C data from stromatolite (II)
14
UCIAMS #
D14C
±
C age
(years BP)
±
Fraction
modern
±
Distance from top of
stromatolite
Lamination
style
48969
48970
48971
48974
48972
48973
)526.0
)706.5
)639.3
)549.4
)806.7
)719.6
0.7
0.6
0.6
1.0
0.8
0.8
5995
9845
8190
6405
13205
10215
15
20
15
15
35
25
0.474
0.293
0.361
0.451
0.193
0.280
0.001
0.001
0.001
0.001
0.001
0.001
0 (top)
1.0 cm
2.5 cm
4.0 cm
5.0 cm
7.0 cm
Drape
Main body
Main body
Drape
Main body
Main body
14
C data – origin of ‘‘drape’’ and ‘‘main body’’ facies
The alternation between the main body and drape style of
lamination suggests that the growth of OPP stromatolites is
related to some environmental forcing mechanism that affects
the microbial community undergoing silicification. Initially,
we assumed that 14C could be used as a complementary dating
tool to the 228Th ⁄ 228Ra, 228Ra ⁄ 226Ra, 137Cs and the growth
experiments. However, the results were quite surprising, and
it is clear that 14C is more useful as a tracer of carbon utilization in the stromatolites rather than their age. Six samples
were analyzed from Stromatolite (II) (Table 2).
Although siliceous stromatolites harbor a robust community of bacteria (Pepe-Ranney et al., 2011), it is unlikely that
the Corg content of the extant microbial population within the
stromatolite influences the 14C data presented here. We offer
the following calculation in support of this statement. Had
there been as many as 107 extant microbes per ml of stromatolite, and assuming a porosity of 80% and a carbon content per
microbe of 60 fg ⁄ cell (Fukuda et al., 1998), the carbon from
this source would amount to <<0.01 wt.% C. The carbon content we measured averaged 1.3 wt.% C (Table S1). Thus, it is
very likely that the Corg content of the stromatolites is a remnant of bacterial sheath carbon as opposed to extant cellular
biomass. Furthermore, the C:N (molar) ratio ranged from 10
to 16 with the average value = 12 (Supplementary Data)
which supports the indigenous origin of the organic matter. If
the source of Corg to this structure were terrestrially derived
(grasses, other higher plants), we would expect to find a much
higher C:N ratio. These results strengthen our assumption
that the organic material preserved within the stromatolite is
primarily associated with the sheath builders entombed in
amorphous SiO2.
14
C ages of various horizons ranged from 5995 to
13,205 years BP and did not occur in stratigraphic order (oldest was not at the base, youngest not at the top), indicating
that radiocarbon is not appropriate as a dating tool for OPP
stromatolites (Table 2). The growth experiments and the
228
Th ⁄ 228Ra, 228Ra ⁄ 226Ra, and 137Cs results are congruent
with one another and suggest that the 14C ages do not represent depositional ages.
14
C is more appropriately used as a tracer of carbon utilization than as a dating tool in OPP. In a hot spring, it seems reasonable to assume that the dissolved inorganic carbon (DIC)
for autotrophic microbial growth originates from two sources:
the CO2-charged vent water and the atmosphere. The inflow
water to Obsidian Pool, which is adjacent to OPP and is well
characterized chemically, has an alkalinity of 2.6 meq ⁄ L
and pH = 6.8 (Shock et al., 2005) suggesting a pCO2 (aq)
of 300–600 lM. Much of this is likely derived from CO2 in
the subsurface spring source waters as Shock et al. (2005)
have documented the bubbles emerging from this spring are
>95% CO2. The original source of the DIC contained in vent
waters is likely radiocarbon dead, originating directly from the
magma in the YNP volcanic system, or perhaps from the
Paleozoic limestones adjacent to the volcanic system. The
atmosphere provides another source for CO2 that will
exchange with the pool waters as the spring waters cool. CO2
derived from the atmosphere will have a modern D14C signature. Cyanobacteria growing in a hot spring will acquire DIC
sourced from a mixture of spring water and the atmosphere,
more from the atmosphere if they are located very near the
water–air interface or emergent, and more from the vent if
they are submerged deeper within the spring.
We constructed a mixing line using standard convention
where 14C-dead carbon = )999 and modern carbon is +100
(Fig. 7). This choice for the value of atmospheric 14CO2 is reasonable for the 1990s – early 2000s (Levin & Kromer, 2004)
yet its value will not affect the outcome by more than a few percent. The fraction of atmospheric CO2 vs. vent CO2 incorpo–400.0
Delta 14C
Stromatolite (II) upper layer, suggesting that Stromatolite
(II) was collected approximately 1 year after it had stopped
growing.
pe
Dra
–600.0
–1000.0
0.00
ody
in b
Ma
–800.0
0.20
0.40
0.60
Fraction of modern carbon
Fig. 7 Mixing line between carbon dead ()999) and modern (+100) carbon,
presented as the fraction of modern carbon. Drape facies consistently records a
greater fraction of modern carbon vs. the main body.
2011 Blackwell Publishing Ltd
Siliceous stromatolite growth rates
Increased depth of submergence likely corresponds with
increased temperature, and near-emergence from the spring
will correspond with cooler temperatures; the fact that diatoms (eukaryotes – more sensitive to higher temperatures than
many cyanobacteria) were only present in the drape facies
corroborates our interpretation of the depth ⁄ temperature
changes. Furthermore, the fact that the drape facies is consistently more completely cemented with silica also corroborates
deposition under a lower temperature or perhaps even emergent. The 14C results, species analyses, and silica cementation
suggest that the drape fabrics formed when the spring was low
and main body fabrics formed when the spring was higher.
rated into stromatolite biomass during photosynthesis is a
function of where it falls along the mixing line. Our data indicate that the carbon acquired and retained within the stromatolite was fixed using DIC that was an admixture of radiocarbon
dead and modern CO2 (Table 2). The proportion of these two
DIC sources varied during growth of this structure, but water
from the vent accounts for 53–81% of the total fixed Corg.
The samples corresponding to the drape fabric record a
greater fraction of atmospheric modern CO2, whereas the
samples from the main body were sourced more from the vent
(Fig. 7). As the stromatolite began to grow, the photosynthetic communities used predominantly vent-derived CO2,
suggesting that the structure was deeper in the pool, or
submerged as growth initiated. About half way through the
stromatolite’s growth, the first drape was deposited. Because
the drape records greater influence from atmospheric CO2,
the structure may have been much closer to the water–air
interface or perhaps even emergent, likely due to a drop in
pool-level relative to the stromatolite. Following the first
drape, our interpretation of the 14C data suggest that the pool
waters deepened followed by growth of more main body
facies. Finally, at the top of the stromatolite, a drape facies
formed with a 14C signature suggesting CO2 sourced equally
from the vent and the atmosphere.
A
419
Environmental forcing of stromatolite lamination
Stromatolite (III) grown on a planted substrate provides both
an understanding of growth rate but also insight into the lamination formation process. The occurrence of 80 couplets in
141 days indicates a laminae frequency of 1 couplet every
1.75 days, which does not immediately suggest an obvious
diurnal, solar or lunar cyclicity. However, it is possible that the
diurnal cycle did dictate growth, but that the structure did not
initiate growth immediately or that the main body ceased
growth prior to collection. To further investigate forcings that
60
Temperature (oC)
55
50
45
40
02/26/11
168
02/06/11
12/28/10
144
01/17/11
12/08/10
11/18/10
10/29/10
10/09/10
09/19/10
08/30/10
08/10/10
07/21/10
07/01/10
06/11/10
35
B
58
Temperature (oC)
Date
56
54
52
50
0
24
48
72
96
120
Hours (7/17–7/23)
Fig. 8 (A) Obsidian Prime Pool temperature from June 2010 through February 2011 (with a few week hiatus in October to turn-around the temperature sensors).
(B) Hourly temperature data from 1 week in July 2010.
2011 Blackwell Publishing Ltd
420
W. M. BERE LS ON et a l.
may occur with a frequency of approximately one couplet
every 1–1.75 days, we placed probes in OPP in the summer of
2010 through the winter 2011 in order to document variability in water temperature. We note that Stromatolite (III) grew
in the fall-winter of 2008–2009 and thus our 2010–2011
temperature measurements may not directly apply to this
structure. Nonetheless, temperature probes do show interesting features (Fig. 8) that may relate to stromatolite laminae
formation.
Interestingly, the high-resolution record of temperature
reveals that, like stromatolite lamination frequency, the record
of temperature excursions in OPP is nearly, but not exactly
diurnal, where weather phenomena are superimposed on the
diurnal temperature cycle. The winter average temperature is
50 ± 4 C, with anomalies associated with storms approaching 15 C. The summer average is 55 C, with weather related
excursions to 46 C. The temperature excursions to lower values occur more often in the winter and are temporally correlated with low pressure and high wind events as recorded at
West Yellowstone Airport. Figure 8B shows an expanded view
of pool temperature fluctuations over 7 days in July 2010.
Only 5 of the 7 days show clear temperature maxima and only
five evenings show clear minima. At Yellowstone, there are
often days where daytime temperatures are not much warmer
than night-time temperatures, even in summer. The lack of a
diurnal temperature signal is also apparent in the winter data,
there appear to be many days when daytime and night-time
water temperatures are similar. This high-resolution temperature record suggests that temperature variability may not
occur on a daily basis.
Temperature is an important parameter to consider as it is
possible that cooling increases supersaturation of SiO2 (amorphous), and this mineralization should help to drive the laminae couplet formation process. We calculated OPP degree of
saturation with respect to amorphous SiO2 (Fig. 9). OPP is
7.0
K or [H4SiO4] in mM
6.0
Range for
Obsidian’
5.0
4.0
Supersaturation
3.0
Undersaturation
2.0
1.0
0
0
20
40
Temperature
60
80
(oC)
Fig. 9 Solubility plot of amorphous SiO2 as a function of temperature (C).
Thermodynamic values obtained from Thermodyn (http://www.microeco.uzh.ch/therm/thermodyn.html). OPP dissolved Si concentrations range
from 4.9 to 5.9 mM and are shown as the gray box.
always supersaturated, and that a 4–15 C cooling causes a
higher degree of supersaturation. Although thermodynamically supersaturated, the process of silica deposition may not
directly respond to thermodynamic rules, hence the temperature forcing of a few degrees (more during winter months),
may be the necessary driver that leads to mineralization (Hinman & Lindstrom, 1996). This process could account for horizontal lamination if precipitation of SiO2 leads to the greater
tendency for sheath tube close packing and horizontal or
sub-horizontal orientation. This study points out the possibility of diurnal or near-diurnal forcing via changes in pool
temperature.
Light intensity is another parameter that may be linked to
laminae couplet formation. We do not have data on this environmental parameter but consider it likely that diurnal fluctuations may be interrupted by days of excessive cloudiness and
low light levels. In this fashion, light intensity, like pool temperature fluctuations, may not follow a strict diurnal cycle, but
it is likely to be near-diurnal.
Environmental forcing of main body–drape facies
Based on the interpretation of the 14C data, we propose that
stromatolite lamination style (main body or drape) provides
a record of changes in spring level in OPP over a period of
less than 10 years. Why did the spring level change – do the
stromatolites constitute high-resolution recorders of some
larger scale process in YNP, perhaps related to the tectonic
activity in the area? The USGS monitors earthquake activity
and ground motion in the YNP region (seismic activity and
volcanic eruptions pose significant potential harm and
destruction to humans in this area). Earthquakes in YNP are
commonly associated with magma movement beneath the
caldera, and such movement may affect local spring level
and ⁄ or spring chemistry. Interestingly, the pattern of earthquake activity (Brantley et al., 2004) reveals cyclicity on several timescales, well within the presumed accumulation rate
of the stromatolites (Fig. 10). In addition, the caldera
undergoes expansion and contraction on 20 year periodicity, which could also affect spring levels. Although we do
not make discrete correlations at this time, we have plotted
the pattern of caldera uplift and subsidence next to spring
level as suggested by the 14C analyses (Fig. 10). The potential that stromatolite growth represent subtle changes in
YNP or local ground motion is a hypothesis we present for
consideration.
Relevance to the ancient rock record
Most ancient stromatolites are found in marine strata and are
composed of calcium carbonate, so it is important to specifically address how the OPP stromatolites – formed in a hot
spring environment and composed of silica-coated filaments –
inform our interpretation of ancient stromatolites. It seems
2011 Blackwell Publishing Ltd
Siliceous stromatolite growth rates
A
0
20
Relative
Spring
Depth
40
Base
Top
Distance from base of strom
1000
Cald
era s
u
(19 m bsidence
m/ye
ar)
ft
ra upli
Calde
500
0
1985
1990
Year
1995
2000
Growth of Strom III
100
2006
2007
Year
2008
50
0
2009
Vertical movement (mm)
C
2005
Earthquakes per quarter
B
60
Shallower
Percent modern 14C
Deeper
421
lead to less distinctly laminated structures. Rare fine laminations do exist in some modern marine stromatolites, and like
the OPP lamination, they likely formed rapidly (relative to the
growth of the stromatolite itself) (e.g., Reid et al., 2000; Visscher et al., 2000). The OPP stromatolites provide a textural
analog to the ancient forms, if not a geochemical analog,
where very early lithification – likely on the timescale of the
diurnal life of a microbial mat – locks in the fine microbial lamination before degradation destroys it. Others have suggested
that carbonate saturation in the oceans may have been much
higher in the past (Kempe & Degens, 1985; Grotzinger,
1990; Grotzinger & Kasting, 1993), a feature that may have
been more important for stromatolite formation than previously appreciated (cf., Grotzinger, 1990).
In general, if it is not possible to date the rate of formation
of ancient stromatolites, the accretion history of ancient stromatolites will remain largely unknown. However, the OPP
stromatolites provide a case study with respect to what is possible with respect to growth rate. On one hand, it has long been
known that many phototrophic microbial mats will respond to
diurnal or nearly diurnal cycles (e.g., like the main body facies
of the OPP stromatolites) (e.g., Walter et al., 1972; Walter,
1976). On the other hand, the OPP stromatolites reveal the
importance of the different lamination styles with respect to
accretion history and timing. Interestingly, the two styles of
lamination present in the OPP stromatolites represent forcing
on two timescales; the finely laminated ‘‘main body’’ represents diurnal or nearly diurnal growth, and the alternation of
‘‘main body’’ and ‘‘drape’’ fabric represents fluctuation in
pool height ⁄ volume occurring on a scale of years. Although it
might not be possible to know the accretion rate of different
lamination styles in ancient stromatolites, the OPP stromatolites inform us that drastic differences in timing of lamination
formation in a single stromatolite is possible. Thus, the linkage
of different lamination styles (main body and drape) to specific
environmental conditions and timescales represents a step in
understanding how to interpret similar changes in ancient
structures where environmental conditions are poorly known.
Fig. 10 (A) Our interpretation of the changes in spring level in OPP, based on
the fraction of modern vs. vent carbon found in the stromatolite. When the
spring level is high (i.e., the stromatolites are deeper, covered by more water),
there is less modern 14C incorporated, and vice versa. (B) We propose that the
stromatolite response to spring level change record some larger scale process in
YNP. Here, we re-plot the USGS record of earthquake activity in YNP and the
overall pattern of caldera uplift and subsidence (modified from Brantley et al.,
2004) presented as candidate processes that might control spring level – both
operate on a reasonable time scale with respect to stromatolite growth,
although earthquake activity is perhaps more in line with the preferred growth
rates (5–8 years with several drapes). (C) Vertical land motion in Hayden Valley
region (near OPP). Data obtained from University of Utah Seismology Research
Group (http://www.uusatrg.utah.edu/index.html).
SUMMARY
clear from the OPP stromatolites that early lithification of
microbial mats is important in the formation of finely laminated structures. Lithification separates mats (not lithified)
from stromatolites (lithified), so it is not surprising that lithification would be important. In OPP, lithification of the main
body facies occurs on the time scale of the life of a biofilm, and
in the case of the light layers, the diurnal life-cycle of a bacterium, where intricate filamentous structures are preserved in
life position before degradation destroys them. Without early
lithification, whether silica or carbonate, it is unlikely that a
finely laminated structure representing an ancient microbial
mat would be preserved; the degradation process would likely
An analysis of siliceous stromatolites from the OPP in Yellowstone National Park has led to some novel interpretations of
their growth rate and lamination style that are likely relevant
to stromatolites throughout the rock record.
(i) There are two distinct growth habits of siliceous stromatolites, (a) a 100–200 lm scale layering of a uniform tubeor sheath-building cyanobacteria such that more and less
dense layers accrete alternatively, (b) layers of the sheath
builders are interrupted by a ‘‘drape’’ style of laminae, lacking
internal structure and composed of more densely silicified bacterial forms and diatoms.
(ii) Three radiometric dating techniques, 228Th ⁄ 228Ra,
228
Ra ⁄ 226Ra, and 137Cs are internally consistent and predict
2011 Blackwell Publishing Ltd
422
W. M. BERE LS ON et a l.
stromatolite accretion of centimeters over the course of years.
These data are consistent with a growth rate experiment in
which 2 cm of stromatolite growth occurred over 5 months.
Main body growth is rapid but is punctuated by periods of
slower or no growth.
(iii) We demonstrate the utility of 14C content as a tracer of
CO2 source utilized by the autotrophic, stromatolite-building
communities. Two end-member sources, vent water and
atmospheric were used to develop a mixing line on which stromatolite samples lie. Portions of the well-laminated main body
have D14C signatures indicating a high proportion of their
CO2 originated from the hot spring vent (65–80%). The
drape fabric of a stromatolite has a D14C signature indicating a
larger fraction of atmospheric CO2 (46%) was fixed by autotrophs living in this layer. Our interpretation of this data is
that main body growth occurred when the spring was deeper
and drape facies develop as the stromatolite is emergent.
(iv) Temperature is likely an important parameter in the
formation of silicious mineralization especially in environments where the water–air differential can exceed 80 C. High
temporal resolution temperature time series data collected
over the summer 2010–winter 2011 indicate 4 C average
day–night differences in pool water temperature with excursions due to high wind events of >15 C. Fluctuations are primarily diurnal in the summer months although there are days
when temperature maxima or minima do not develop. In the
winter months, variability between day ⁄ night is more irregular, thus temperature forcing of silica deposition may not
record diurnal cycles exactly.
We conclude that two styles of lamination represent forcing
on two timescales; the finely laminated ‘‘main body’’ represents diurnal or nearly diurnal growth and the alternation of
‘‘main body’’ and ‘‘drape’’ fabric represents fluctuation in
pool height ⁄ volume occurring on a scale of years. This study
adds strength to the argument that stromatolites can be
valuable records of microbial community structure and
environmental forcing. The combination of geochemical,
high-resolution environmental monitoring and microbiological ⁄ molecular analyses will help further our understanding of
these modern analogs of ancient life-forms.
ACKNOWLEDGMENTS
We thank the Yellowstone Center for Resources and Christie
Hendrix for assistance with a scientific research and collecting
permit to J.R.S. Funding for this work was provided by a
National Science Foundation Microbial Biology Postdoctoral
Start-up Award to J.R.S., and by a U.S. Air Force Office of Scientific Research award to J.R.S. Tony Kampf (Los Angeles
County Museum of Natural History) assisted with XRD analysis, Nick Rollins assisted with the ibutton research. Shannon
Ulrich, Lee Kump and anonymous reviewers provided useful
critiques and feedback on versions of this manuscript. We
acknowledge the support of the Agouron Institute, NASA
Exobiology, the GB Moore Foundation and credit the International GeoBiology Summer Course, students (in particular,
Scott Mata and Cara Harwood) and instructors (Kurt Hanselmann) for several years of work and collaboration on this
study. Ann Close and Sue Anderson from the University of
Southern California provided expedition support.
REFERENCES
Allewalt JP, Bateson MM, Revsbech NP, Slack K, Ward DM
(2006) Effect of temperature and light on growth of and photosynthesis by Synechococcus isolates typical of those predominating in the Octopus Spring microbial mat community of
Yellowstone National Park. Applied and Environmental Microbiology, 72, 544–550.
Allwood AC, Walter MR, Kamber BS, Marshall CP, Burch IW (2006)
Stromatolite reef from the Early Archaean era of Australia. Nature,
114, 714–718.
Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ (1990) Basic
local alignment search tool. Journal of Molecular Biology, 215,
403–410.
Amann RI, Ludwig W, Schleifer KH (1995) Phylogenetic identification and in situ detection of individual microbial cells without cultivation. Microbiological Reviews, 59, 143–169.
Awramik SM, Grey K (2005) Stromatolites: biogenicity, biosignatures, and bioconfusion. Proceedings of SPIE, 5906: 5906P1–5906P-9.
Awramik SM, Riding RE (1988) Role of algal eukaryote in subtidal
columnar stromatolite formation. Proceedings of the National
Academy of Sciences (USA), 85, 1327–1329.
Awramik SM, Vanyo JP (1986) Heliotropism in modern stromatolites. Science, 231, 1279–1281.
Barghoorn ES, Tyler SA (1965) Microorganisms from the Gunflint
Chert: these structurally preserved Precambrian fossils from
Ontario are the most ancient organisms known. Science, 147, 563–
575.
Benning LG, Phoenix VR, Mountain BW (2005) Biosilicification: the
role of cyanobacteria in silica sinter deposition. In Micro-Organisms
and Earth Systems: Advances in Geomicrobiology (eds Gadd MG,
Semple TK, Lappin-Scott MH). Society for General Microbiology
Symposium, Cambridge University Press, Cambridge, pp. 131–
150.
Bhaya D, Grossman AR, Steunou A-S, Khuri N, Cohan FM, Hamamura N, Melendrez MC, Bateson MM, Ward DM, Heidelberg JF
(2007) Population level functional diversity in a microbial community revealed by comparative genomic and metagenomic analyses.
The ISME Journal, Multidisciplinary Journal of Microbial Ecology,
1, 703–713.
Blank CE, Cady SL, Pace NR (2002) Microbial composition of nearboiling silica-depositing thermal springs throughout Yellowstone
National Park. Applied and Environmental Microbiology, 68,
5123–5135.
Bosak T, Bush J, Flynn M, Liang B, Ono S, Petroff AP, Sim MS
(2010) Formation and stability of oxygen-rich bubbles that shape
photosynthetic mats. Geobiology, 8, 1–11.
Brantley SR, Lowenstern JB, Christiansen RL, Smith RB, Heasler H,
Waite G, Wicksthermal C (2004) Tracking changes in Yellowstone’s restless volcanic system. US Geological Survey Fact Sheet
100-03, 4.
Bryant DA, Costas AMG, Maresca JA, Chew AGM, Klatt CG, Bateson MM, Tallon LJ, Hostetler J, Nelson WC, Heidelberg JF, Ward
DM (2007) Candidatus Chloracidobacterium thermophilum: an
2011 Blackwell Publishing Ltd
Siliceous stromatolite growth rates
aerobic phototrophic acidobacterium. Science, 317, 523–
526.
Cady SL, Farmer JD (1996) Fossilization processes in siliceous thermal springs: trends in preservation along thermal gradients. In Evolution of Hydrothermal Ecosystems on Earth (and Mars?), Ciba
Foundation Symposium 202. Wiley, Chichester, pp. 150–173.
Caporaso JG, Kuczynski J, Stombaugh J, Bittinger K, Bushman FD,
Costello EK, Fierer N, Gonzalez Peña A, Goodrich JK, Gordon
JI, Huttley GA, Kelley ST, Knights D, Koenig JE, Ley RE,
Lozupone CA, McDonald D, Muegge BD, Pirrung M, Reeder J,
Sevinsky JR, Turnbaugh PJ, Walters WA, Widmann J, Yatsunenko
T, Zaneveld J, Knight R (2010) QIIME allows analysis of
high-throughput community sequencing data. Nature Methods,
7, 335–336.
Chivas AR, Torgersen T, Polach HA (1990) Growth rates and
Holocene development of stromatolites from Shark Bay, Western
Australia. Australian Journal of Earth Sciences, 37, 113–121.
Dill RF, Shinn EA, Jones AT, Kelly K, Steinen RP (1986) Giant subtidal stromatolites forming in normal salinity waters. Nature, 324,
55–58.
Doemel WN, Brock TD (1974) Bacterial stromatolites: origin of laminations. Science, 184, 1083–1085.
Dravis JJ (1983) Hardened subtidal stromatolites, Bahamas. Science,
219, 385–386.
Dupraz C, Patissina R, Verrecchia EP (2006) Simulation of stromatolite morphospace using ‘DLA-CA’ growth model’: translation of
energy in morphology. Sedimentary Geology, 185, 185–203.
Edgar RC (2010) Search and clustering orders of magnitude faster
than BLAST. Bioinformatics, 26, 2460–2461.
Ferris MJ, Kuhl M, Wieland A, Ward DM (2003) Cyanobacterial ecotypes in different optical microenvironments of a 68 Degree C Hot
Spring Mat Community revealed by 16S-23S rRNA internal transcribed spacer region variation. Applied and Environmental Microbiology, 69, 2893–2898.
Finsinger K, Scholz I, Serrano A, Morales S, Uribe-Lorio L, Mora M,
Sittenfeld A, Weckesser J, Hess WR (2008) Characterization of
true-branching cyanobacteria from geothermal sites and hot springs
of Costa Rica. Environmental Microbiology, 10, 460–473.
Fukuda R, Ogawa H, Nagata T, Koike I (1998) Direct determination
of carbon and nitrogen contents of natural bacterial assemblages in
marine environments. Applied and Environmental Microbiology,
64, 3352–3358.
Grotzinger JP (1990) Geochemical model for Proterozoic stromatolite decline. American Journal of Science, 290-A, 80–103.
Grotzinger JP, Kasting JF (1993) New constraints on Precambrian
ocean composition. Journal of Geology, 101, 235–243.
Grotzinger JP, Knoll AH (1999) Stromatolites in Precambrian carbonates; evolutionary mileposts or environmental dipsticks?
Annual Review of Earth and Planetary Sciences, 27, 313–358.
Grotzinger JP, Rothman DH (1996) An abiotic model for stromatolite morphogenesis. Nature, 383, 423–425.
Guidry SA, Chafetz HS (2003) Depositional facies and diagenetic
alteration in a relict siliceous hot-spring accumulation: examples
from Yellowstone National Park, U.S.A. Journal of Sedimentary
Research, 73, 806–823.
Hamady M, Walker JJ, Harris JK, Gold NJ, Knight R (2008) Errorcorrecting barcoded primers for pyrosequencing hundreds of samples in multiplex. Nature Methods, 5, 235–237.
Hinman NW, Lindstrom RF (1996) Seasonal changes in silica
deposition in hot springs systems. Chemical Geology, 132, 237–
246.
Hoffman P (1976) Stromatolite morphogenesis in Shark Bay, Western Australia. In Stromatolites (ed Walter MR). Elsevier Sci. Publ.
Co., Amsterdam, Netherlands, pp. 261–271.
2011 Blackwell Publishing Ltd
423
Hofmann HJ, Grey K, Hickman AH, Thorpe RI (1999) Origin of
3.45 Ga coniform stromatolites in Warrawoona Group, Western
Australia. Geological Society of America Bulletin, 111, 1256–
1262.
Hugenholtz P, Pitulle C, Hershberger KL, Pace NR (1998) Novel
division level bacterial diversity in a Yellowstone hot spring. Journal
of Bacteriology 180, 366–376.
Hunter J (2007) Matplotlib: a 2D graphics environment. Computing
in Science & Engineering, 9, 90–95.
Kempe S, Degens ET (1985) An early soda ocean. Chemical Geology,
53, 95–108.
Klatt CG, Bryant DA, Ward DM (2007) Comparative genomics
provides evidence for the 3-hydroxypropionate autotrophic
pathway in filamentous anoxygenic phototrophic bacteria and in
hot spring microbial mats. Environmental Microbiology, 9,
2067–2078.
Konhauser KO, Jones B, Reysenbach AL, Renaut RW (2003) Hot
spring sinters: keys to understanding Earth’s earliest life forms.
Canadian Journal of Earth Sciences, 40, 1713–1724.
Konhauser K, Jones B, Phoenix V, Ferris G, Renaut R (2004)
The microbial role in hot spring silicification. Ambio, 33,
552–558.
Lane D (1991) 16S ⁄ 23S rRNA sequencing. In Nucleic Acid Techniques in Bacteria Systematics (eds Stackebrandt E, Goodfellow M).
Wiley and Sons, Chichester, UK, p. 329.
Levin I, Kromer B (2004) The tropospheric CO2 level in mid-latitudes of the Northern Hemisphere (1959–2003). Radiocarbon,
46, 1261–1272.
Logan BW (1961) Cryptozoon and associate stromatolites from the
Recent, Shark bay, Western Australia. Journal of Geology, 69, 517–
533.
Lowe DR (1994) Abiological origin of described stromatolites older
than 3.2 Ga. Geology, 22, 387–390.
McLoughlin NA, Wilson LA , Brasier MD (2008) Growth of synthetic stromatolites and wrinkle structures in the absence of
microbes; implications for the early fossil record. Geobiology, 6,
95–105.
van der Meer MTJ, Schouten S, Sinninghe-Damsté JS, Ward DM
(2007) Impact of carbon metabolism on 13C signatures of
cyanobacteria and green non-sulfur-like bacteria inhabiting a
microbial mat from an alkaline siliceous hot spring in
Yellowstone National Park (USA). Environmental Microbiology, 9,
482–491.
Pepe-Ranney C, Berelson WM, Corsetti FA, Treants M, Spear JR
(2011) The geobiology of stromatolite construction by cyanobacteria in Yellowstone National Park. (Accepted: Environmental
Microbiology and Env. Micro. Reports).
Petryshyn VA, Corsetti FA, Berelson WM, Lund SP, Beaumont W
(2011) Stromatolite lamination frequency; Walker Lake, Nevada.
Implications for stromatolites as biosignatures.
Reeder J, Knight R (2010) Rapidly denoising pyrosequencing amplicon reads by exploiting rank-abundance distributions. Nature
Methods 7, 668–669.
Reid RP, Visscher PT, Decho AW, Stolz JF, Bebout BM, Dupraz C,
Macintyre IG, Paerl HW, Pinckney JL, Prufert-Bebout L, Steppe
TF, Des Marais DJ (2000) The role of microbes in accretion, lamination and early lithification of modern marine stromatolites. Nature, 406, 989–992.
Riding RE (1990) Calcareous Algae and Stromatolites. SpringerVerlag, New York.
Riding RE, Awramik SM (2000) Microbial Sediments. SpringerVerlag, New York.
Schouten S, van der Meer MTJ, Hopmans EC, Rijpstra WIC, Reysenbach A-L, Ward DM, Sinninghe Damste JS (2007) Archaeal and
424
W. M. BERE LS ON et a l.
bacterial glycerol dialkyl glycerol tetraether lipids in hot springs of
Yellowstone National Park. Applied and Environmental Microbiology, 73, 6181–6191.
Shapiro RS (2000) A comment on the systematic confusion of
thrombolites. Palaios, 15, 166–169.
Shock EL, Holland M, Meyer-Dombard DR, Amend JP (2005) Geochemical sources of energy for microbial metabolism in hydrothermal ecosystems: Obsidian Pool, Yellowstone National Park, USA.
In Geothermal Biology and Geochemistry of Yellowstone National
Park (eds Inskeep WP, McDermott TR). Montana State University
Publications, Bozeman, Montana, pp. 95–112.
Spear JR, Walker JJ, McCollom TM, Pace NR (2005) Hydrogen and
bioenergetics in the Yellowstone geothermal ecosystem. Proceedings of the National Academy of Sciences (USA), 102, 2555–2560.
Steunou A-S, Bhaya D, Bateson MM, Melendrez MC, Ward DM,
Brecht E, Peters JW, Kühl M, Grossman AR (2006) In situ analysis
of nitrogen fixation and metabolic switching in unicellular
thermophilic cyanobacteria inhabiting hot spring microbial mats.
Proceedings of the National Academy of Sciences (USA), 103,
2398–2403.
Steunou A-S, Jensen SI, Brecht E, Becraft ED, Bateson MM, Kilian
O, Bhaya D, Ward DM, Peters JW, Grossman AR, Kuhl M (2008)
Regulation of nif gene expression and the energetics of N2
fixation over the diel cycle in a hot spring microbial mat. The ISME
Journal, Multidisciplinary Journal of Microbial Ecology, 2,
364–378.
Stuiver M, Polach HA (1977) Discussion: reporting of 14C data.
Radiocarbon, 19, 355–363.
Visscher PT, Reid RP, Bebout BM (2000) Microscale observations of
sulfate reduction: correlation of microbial activity with lithified
micritic laminae in modern marine stromatolites. Geology 28,
919–922.
Walter MR (1976) Stromatolites. In Developments in Sedimentology
20 (ed. Walter MR). Elsevier, Amsterdam, pp. 790.
Walter MR, Bauld J, Brock TD (1972) Siliceous algal and bacterial
stromatolites in hot spring and geyser effluents of Yellowstone
National Park. Science, 178, 402–405.
Weed WH (1889) US Geological Survey Report 9 (1887–1888).
613.
Wilmotte A, Van der Auwera G, De Wachter R (1993) Structure
of the 16 S ribosomal RNA of the thermophilic cyanobacterium
Chlorogloeopsis HTF (‘Mastigocladus laminosus HTF’) strain
PCC7518, and phylogenetic analysis. FEBS Letters, 317, 96–
100.
SUPPORTING INFORMATION
Additional Supporting Information may be found in the
online version of this article:
Table S1. Dry weight percent C, N, and S from regions of Stromatolite (I) as
indicated by the letters in Fig. 2. Sample I is an integrated sample including
regions A–H.
Please note: Wiley–Blackwell are not responsible for the
content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the
article.
2011 Blackwell Publishing Ltd