Actinobacterial Diversity in Volcanic Caves and Associated Geomicrobiological Interactions

Abstract

Volcanic caves are filled with colorful microbial mats on the walls and ceilings. These volcanic caves are found worldwide, and studies are finding vast bacteria diversity within these caves. One group of bacteria that can be abundant in volcanic caves, as well as other caves, is Actinobacteria. As Actinobacteria are valued for their ability to produce a variety of secondary metabolites, rare and novel Actinobacteria are being sought in underexplored environments. The abundance of novel Actinobacteria in volcanic caves makes this environment an excellent location to study these bacteria. Scanning electron microscopy (SEM) from several volcanic caves worldwide revealed diversity in the morphologies present. Spores, coccoid, and filamentous cells, many with hair-like or knobby extensions, were some of the microbial structures observed within the microbial mat samples. In addition, the SEM study pointed out that these features figure prominently in both constructive and destructive mineral processes. To further investigate this diversity, we conducted both Sanger sequencing and 454 pyrosequencing of the Actinobacteria in volcanic caves from four locations, two islands in the Azores, Portugal, and Hawai'i and New Mexico, USA. This comparison represents one of the largest sequencing efforts of Actinobacteria in volcanic caves to date. The diversity was shown to be dominated by Actinomycetales, but also included several newly described orders, such as Euzebyales, and Gaiellales. Sixty-two percent of the clones from the four locations shared less than 97% similarity to known sequences, and nearly 71% of the clones were singletons, supporting the commonly held belief that volcanic caves are an untapped resource for novel and rare Actinobacteria. The amplicon libraries depicted a wider view of the microbial diversity in Azorean volcanic caves revealing three additional orders, Rubrobacterales, Solirubrobacterales, and Coriobacteriales. Studies of microbial ecology in volcanic caves are still very limited. To rectify this deficiency, the results from our study help fill in the gaps in our knowledge of actinobacterial diversity and their potential roles in the volcanic cave ecosystems.

Full text

Actinobacterial diversity in volcanic
caves and associated geomicrobiological
interactions
Cristina Riquelme1, Jennifer J. Marshall Hathaway2, Maria D. Enes Dapkevicius1, Ana Z. Miller3,
Ara Kooser2, Diana E. Northup2, Valme Jurdo3, Octavio Fernandaz4, Cesareo Saiz-Jimenez3,
Naowarat Cheeptham5*
1Departamento de Ciências Agrárias, Universidade dos Açores, Portugal, 2Department of Biology,
University of New Mexico, USA, 3Consejo Superior de Investigaciones Científicas (IRNAS-CSIC), Instituto
de Recursos Naturales y Agrobiología, Spain, 4Grupo de Espeleología Tebexcorade-La Palma, Spain,
5Biological Sciences, Faculty of Science, Thompson Rivers University, Canada
Submitted to Journal:
Frontiers in Microbiology
Specialty Section:
Extreme Microbiology
Article type:
Original Research Article
Manuscript ID:
162211
Received on:
17 Jul 2015
Revised on:
29 Oct 2015
Frontiers website link:
www.frontiersin.org
In review

Conflict of interest statement
The authors declare that the research was conducted in the absence of any commercial or financial
relationships that could be construed as a potential conflict of interest
Keywords
Actinobacteria, Volcanic Laval Tubes Caves, microbe-mineral interactions, microbial diversity., Metagenomics
Abstract
Word count: 322
Volcanic caves are filled with colorful microbial mats on the walls and ceilings. These volcanic caves are found worldwide, and
studies are finding vast bacteria diversity within these caves. One group of bacteria that can be abundant in volcanic caves, as
well as other caves, is Actinobacteria. As Actinobacteria are valued for their ability to produce a variety of secondary metabolites,
rare and novel Actinobacteria are being sought in underexplored environments. The abundance of novel Actinobacteria in volcanic
caves makes this environment an excellent location to study these bacteria. Scanning electron microscopy (SEM) from several
volcanic caves worldwide revealed diversity in the morphologies present. Spores, coccoid and filamentous cells, many with
hair-like or knobby extensions, were some of the microbial structures observed within the microbial mat samples. In addition, the
SEM study pointed out that these features figure prominently in both constructive and destructive mineral processes. To further
investigate this diversity, we conducted both Sanger sequencing and 454 pyrosequencing of the Actinobacteria in volcanic caves
from four locations, two islands in the Azores, Portugal and Hawai`i and New Mexico, USA. This comparison represents one of the
largest sequencing efforts of Actinobacteria in volcanic caves to date. The diversity was shown to be dominated by
Actinomycetales, but also included several newly described orders, such as Euzebyales, and Gaiellales. Sixty-two percent of the
clones from the four locations shared less than 97% similarity to known sequences, and nearly 71% of the clones were singletons,
supporting the commonly held belief that volcanic caves are an untapped resource for novel and rare Actinobacteria. The amplicon
libraries depicted a wider view of the microbial diversity in Azorean volcanic caves revealing three additional orders,
Rubrobacterales, Solirubrobacterales and Coriobacteriales. Studies of microbial ecology in volcanic caves are still very limited. To
rectify this deficiency, the results from our study help fill in the gaps in our knowledge of actinobacterial diversity and their
potential roles in the volcanic cave ecosystems.
Ethics statement
(Authors are required to state the ethical considerations of their study in the manuscript including for cases
where the study was exempt from ethical approval procedures.)
Did the study presented in the manuscript involve human or animal subjects: No
In review

1
Actinobacterial diversity in volcanic caves and associated
1
geomicrobiological interactions
2
1Cristina Riquelme#, 2Jennifer J. Marshall Hathaway#, 1Maria de Lurdes Nunes
3
Enes Dapkevicius, 3Ana Z. Miller, 2Ara Kooser, 2Diana E. Northup, 3Valme
4
Jurado, 4Octavio Fernandez, 3Cesareo Saiz-Jimenez, and *5Naowarat
5
Cheeptham
6
1Departamento de Ciências Agrárias, Universidade dos Açores, São Pedro, Portugal
7
2Department of Biology, University of New Mexico, Albuquerque, New Mexico, USA
8
3Instituto de Recursos Naturales y Agrobiología, Consejo Superior de Investigaciones
9
Científicas (IRNAS-CSIC), Sevilla, Spain
10
4Grupo de Espeleología Tebexcorade-La Palma, La Palma, Spain
11
5Department of Biological Sciences, Faculty of Science, Thompson Rivers University,
12
Kamloops, BC, Canada
13
14
#These authors have contributed equally to this study
15
*Corresponding Author
16
Naowarat Cheeptham (Ann), Ph.D.
17
Associate Professor of Microbiology
18
Department of Biological Sciences
19
Faculty of Science
20
Thompson Rivers University
21
900 McGill Road
22
Kamloops, BC
23
V2C 0C8
24
Email: ncheeptham@tru.ca
25
26
27
28
29
30
31
In review

2
ABSTRACT
32
Volcanic caves are filled with colorful microbial mats on the walls and
33
ceilings. These volcanic caves are found worldwide, and studies are finding
34
vast bacteria diversity within these caves. One group of bacteria that can be
35
abundant in volcanic caves, as well as other caves, is Actinobacteria. As
36
Actinobacteria are valued for their ability to produce a variety of secondary
37
metabolites, rare and novel Actinobacteria are being sought in underexplored
38
environments. The abundance of novel Actinobacteria in volcanic caves makes
39
this environment an excellent location to study these bacteria. Scanning
40
electron microscopy (SEM) from several volcanic caves worldwide revealed
41
diversity in the morphologies present. Spores, coccoid and filamentous cells,
42
many with hair-like or knobby extensions, were some of the microbial
43
structures observed within the microbial mat samples. In addition, the SEM
44
study pointed out that these features figure prominently in both constructive
45
and destructive mineral processes. To further investigate this diversity, we
46
conducted both Sanger sequencing and 454 pyrosequencing of the
47
Actinobacteria in volcanic caves from four locations, two islands in the
48
Azores, Portugal and Hawai`i and New Mexico, USA. This comparison
49
represents one of the largest sequencing efforts of Actinobacteria in volcanic
50
caves to date. The diversity was shown to be dominated by Actinomycetales,
51
but also included several newly described orders, such as Euzebyales, and
52
Gaiellales. Sixty-two percent of the clones from the four locations shared less
53
than 97% similarity to known sequences, and nearly 71% of the clones were
54
In review

3
singletons, supporting the commonly held belief that volcanic caves are an
55
untapped resource for novel and rare Actinobacteria. The amplicon libraries
56
depicted a wider view of the microbial diversity in Azorean volcanic caves
57
revealing three additional orders, Rubrobacterales, Solirubrobacterales and
58
Coriobacteriales. Studies of microbial ecology in volcanic caves are still very
59
limited. To rectify this deficiency, the results from our study help fill in the
60
gaps in our knowledge of actinobacterial diversity and their potential roles in
61
the volcanic cave ecosystems.
62
63
INTRODUCTION
64
Actinobacteria are an ubiquitous phyla found to thrive in almost any
65
environment, from soil and marine, to less expected environments such as
66
insects, plants, roots, and caves (See Tiwari and Gupta, 2013 and Subramani
67
and Aalbersberg, 2013 for reviews). Recent culture independent studies have
68
found Actinobacteria in high abundance in a variety of cave types, including
69
volcanic caves (Pašić et al., 2010; Northup et al., 2011; Cuezva., et al. 2012;
70
Niyomyong et al., 2012; Quintana et al., 2013; Barton et al., 2014; Hathaway
71
et al., 2014). Furthermore many characterized species of Actinobacteria have
72
been described from caves (Groth et al., 1999; Lee et al., 2000, 2001, 2006;
73
Jurado et al., 2005 a, b).
74
Primary and secondary metabolites from Actinobacteria have been
75
described as important sources of industrial compounds (Miao and Davies,
76
2010). Rare Actinobacteria, important for novel secondary metabolite
77
production, have been found in many different soil types (Tiwari and Gupta,
78
In review

4
2012; Guo et al., 2015), but caves, volcanic caves included, remain an
79
underexploited environment to screen for industrially important compounds.
80
Goodfellow and Fiedler (2010) suggested examining underexploited sources
81
of Actinobacteria and using taxonomic diversity as a surrogate for chemical
82
diversity, based on the assumption that novel species may contain unique
83
compounds, reducing the re-discovery of the same handful of known
84
secondary metabolites.
85
Cave Actinobacteria are of particular interest because of the unique
86
environment in which they live. The extreme (i.e. low nutrient inputs, low
87
productivity) and often pristine environment would result in bacteria
88
exploiting different metabolic pathways, including the capacity for
89
biomineralization and rock-weathering (Cuezva et al., 2012; Miller et al.,
90
2012a, b). Caves are characterized by microenvironments, which result from
91
several types of reactions, including microbial processes that often involve
92
redox reactions (Barton and Northup, 2007). These mineral microniches
93
control the diversity of subsurface microbial populations (Jones and Bennett,
94
2014), since microbial colonization of rock surfaces is driven by the rock’s
95
chemistry and the organism’s metabolic requirements and tolerances,
96
suggesting that subsurface microbial communities have specific associations to
97
specific minerals. In fact, caves on Earth can harbor a wide variety of mineral-
98
utilizing microorganisms that figure prominently in the formation of secondary
99
mineral deposits and unusual mineralized microstructures recognized as
100
biosignatures. Tubular mineralized sheaths (Boston et al., 2001; Northup et al.,
101
In review

5
2011), bacteria concealed within mineral deposits (Northup et al., 2011),
102
microfossils preserved in minerals (Provencio and Polyak, 2010; Souza-
103
Egipsy et al., 2010), filamentous fabrics (Hofmann et al., 2008) and “cell-
104
sized” etch pits or microborings produced by bacteria (McLoughlin et al.,
105
2007) are some of the proposed models for biosignatures found in subsurface
106
environments.
107
The main goal of the research presented here is to obtain a better
108
understanding of the actinobacterial diversity in volcanic caves from different
109
parts of the world. Comprehensive studies on microbial community ecology of
110
caves identifying abundant, rare and novel species and their environmental
111
implications are still scarce. In the course of this study, we aim to unravel the
112
diversity and composition of volcanic cave Actinobacteria, some of the
113
biogeochemical role of Actinobacteria in caves and their geomicrobiological
114
interactions. Recently, a rapid expansion of interest in subsurface
115
environments has emerged to better understand biodiversity, origins of life on
116
Earth and on other planets. In fact, the reported early results on liquid water
117
and rather recent volcanic activity yielding volcanic caves on Mars, suggesting
118
that the Martian subsurface can house organic molecules or traces of microbial
119
life (Léveillé and Datta, 2010; Northup et al. 2011), make the search for
120
microbial life on Earth’s volcanic caves even more compelling. Overall, this
121
work helps us to understand whether volcanic caves under study present
122
similar levels of diversity and do Actinobacteria found in volcanic caves show
123
In review

6
diversity across different scales from community level to morphology to
124
microbe-mineral interactions.
125
126
127
MATERIALS AND METHODS
128
Morphological characterization of colored microbial mats
129
Sampling of Azorean, Canadian, Canarian, Hawaiian and New Mexican
130
volcanic caves
131
Samples of visible white and/or yellow microbial mats on volcanic
132
cave walls and ceilings (Figure 1) were collected from: 1) Gruta de Terra Mole
133
and Gruta dos Montanheiros in Terceira and Pico Islands, Azores (Portugal);
134
2) Helmcken Falls Cave, British Columbia (Canada); 3) Fuente de la Canaria,
135
Falda de La Horqueta, Llano de los Caños and Honda del Bejenado caves in
136
La Palma Island, Canary islands (Spain), 4) Bird Park Cave and Kipuka
137
Kanohina Cave System, Hawai'i (USA), and 5) Cave 12 from El Malpais
138
National Monument, New Mexico (USA). Samples were taken by gently
139
scraping the colored microbial mats with a sterile scalpel, gathering it into
140
sterile vials and stored at 4ºC until laboratory procedures.
141
Scanning electron microscopy
142
Bulk samples with microbial mats from Canarian volcanic caves
143
(Spain) were directly mounted on a sample stub and sputter coated with a thin
144
gold/palladium film. Samples were subsequently examined on a Jeol JSM-
145
7001F field emission scanning electron microscope (FESEM) equipped with
146
In review

7
an Oxford X-ray energy dispersive spectroscopy (EDS) detector. FESEM
147
examinations were operated in secondary electron (SE) detection mode with
148
an acceleration potential of 15 kV at Instituto Superior Tecnico, University of
149
Lisbon, Portugal. Samples from Helmcken Falls Cave (Canada) were
150
prepared, processed, and observed at the University of British Columbia
151
(UBC) BioImaging Facility (Cheeptham et al., 2013). Rock chips with
152
microbial mats from Azores, New Mexico and Hawai'i were mounted,
153
processed and observed as described in Hathaway et al. (2014).
154
155
Estimation, description and novelty of actinobacterial diversity
156
Sample collection and clone library preparation and OTU-based analysis
157
for New Mexico (USA), Hawai'i (USA), and Azores islands (Portugal)
158
Microbial mat samples of various colors were collected from the dark zone of five
159
caves (Cave 12, Cave 255, Cave 266, Cave 261 and Cave 315) from El Malpais
160
National Monument, New Mexico, six caves on the Big Island of Hawai'i (Bird Park,
161
Epperson’s, Kaumana, and Thurston Caves and the Maelstrom and Kula Kai Caverns
162
Sections of the Kipuka Kanohina Cave System), four caves on the Azorean island of
163
Pico (Furna do Lemos, Gruta dos Montanheiros, Gruta da Ribeira do Fundo and Gruta
164
das Torres) and 11 caves on the Azorean island of Terceira (Algar do Carvão, Gruta das
165
Agulhas, Gruta da Achada, Gruta dos Buracos, Gruta dos Balcões, Gruta da Branca
166
Opala, Gruta da Madre de Deus, Gruta do Natal, Gruta da Terra Mole, Gruta dos
167
Principiantes and Gruta da Malha), see Figure 1 and Supplemental Table 1. DNA from
168
microbial mats of various colors was aseptically collected. DNA was extracted and
169
In review

8
purified using the MoBio PowerSoilTM DNA Isolation Kit using the manufacturer’s
170
protocol (MoBio, Carlsbad, CA), with the exception of the substitution of bead beating
171
for 1.5 min (Biospec Products, Bartlesville, OK, USA) instead of vortexing for cell
172
lysis. 16S rDNA sequences were amplified with universal bacterial primers 46 forward
173
(5′ -GCYTAAYACATGCAAGTCG- 3′) and 1409 reverse (5′ -
174
GTGACGGGCRGTGTGTRCAA- 3′) (Northup et al. 2010).
175
Amplification reactions were carried out in a 25-μL volume with 1X PCR buffer with
176
1.5 mM Mg2+, 0.4 μM of each primer, 0.25 mM of each dNTPs, 5 μg of 50 mg/mL
177
BSA (Ambion, Austin, TX, USA) and 1U AmpliTaq LD (Applied Biosystems, Foster
178
City, CA, USA), and carried out under the following thermocyling conditions on an
179
Eppendforf Mastercycler 5333 (Eppendorf, Hauppauge, NY, USA): 94◦C for 5 min,
180
followed by 31 cycles of 94◦C for 30 sec, 50◦C for 30 sec, 72◦C for 1.5 min, with a
181
final extension at 72◦C for 7 min. Amplicons were cleaned and purified using the
182
Qiagen PCR cleanup kit (Qiagen, Germantown, Maryland) and cloned using the TOPO
183
TA Cloning kit (Invitrogen, Carlsbad, CA). Sequencing was carried out at the
184
Washington University Genome Sequencing Facility. The subset of Actinobacteria
185
were identified with RDP classifier (Maidak et al., 2001), and used for further analysis.
186
Alignments of the resulting actinobacterial sequences set were
187
generated using INFERNAL (Nawrocki et al., 2009), trimmed to 104 to 1403
188
bp to remove ragged ends, and clustered into Operational Taxonomic Units
189
(OTUs) at 97% similarity with QIIME using uclust (Caporaso et al., 2010).
190
In review

9
Taxonomy was assigned using uclust against the greengenes 13.8 database
191
(Edgar, 2010, MacDonald et al., 2012). Sequences were compared with the
192
GenBank database in March 2015 using the Basic Local Alignment Search
193
Tool (BLAST)
1
to determine closest relatives (Altschul et al., 1997). An
194
identity matrix was generated using Bio Edit
2
. The tree was built using
195
FastTree with the gamma and nt options (Price et al., 2009, 2010). OTUs and
196
location were added to the tree using the phyloseq package in R (McMurdie
197
and Holmes, 2013; R Core Team, 2015).
198
All other OTU-based approaches were performed with software
199
package mothur 1.34 (Schloss et al., 2009). Rarefaction curves, nonparametric
200
diversity indexes npsShannon (Chao and Shen, 2003), Shannon (Shannon,
201
1948) and Simpson (Simpson, 1949) and estimator Chao1 (Chao, 1984), as
202
well as the Good’s Coverage (Good, 1953) were calculated to infer the
203
richness and evenness of the samples.
204
205
16S rRNA gene amplicon library preparation, pyrosequencing,
206
bioinformatics and OTU-based analysis in Azorean volcanic caves
207
16S rRNA gene amplicon libraries were prepared from the previously
208
described Azorean microbial mat samples collected from the previously
209
mentioned caves with the exception of Algar do Carvão (Supplemental Table
210
1
www.ncbi.nlm.nih.gov/BLAST/
2
www.mbio.ncsu.edu/BioEdit/ bioedit. html
In review

10
1). The small subunit rRNA gene was amplified from community DNA
211
targeting the V1 and V3 hypervariable region, with barcoded fusion primers
212
containing the Roche-454 A and B Titanium sequencing adapters, a eight-base
213
barcode sequence, the universal forward primer 5’–
214
AGRGTTTGATCMTGGCTCAG -3’ and the universal reverse primer 5’–
215
GTNTTACNGCGGCKGCTG-3’. Amplicon 454 pyrosequencing, as
216
originally described by Dowd et al. (2008), was performed with PCR
217
amplification as described in Brantner et al. (2014). Following PCR, all
218
amplicon products from different samples were mixed in equal concentrations
219
and purified using Agencourt Ampure beads (Agencourt Bioscience
220
Corporation, MA, USA). Samples were sequenced utilizing Roche 454 FLX
221
titanium instruments and reagents and following manufacturer’s guidelines.
222
The raw pyrosequencing reads were processed using version 1.34 of
223
the mothur software package (Schloss et al., 2009). Sequencing reads were
224
assigned to the appropriate samples based on the corresponding barcode and
225
were quality filtered to minimize the effects of random sequencing errors, by
226
eliminating sequence reads <200 bp, sequences that contained more than one
227
undetermined nucleotide (N) and sequences with a maximum homopolymer
228
length of 8 nucleotides. Identification and removal of chimeras was performed
229
with Chimera.uchime (Schloss et al., 2011). Sequences not passing these
230
In review

11
quality controls were discarded. When preparing the inputs for analysis, the
231
‘remove.groups’ command was used to discard all sequences not belonging to
232
the phyla Actinobacteria.
233
OTUs were assigned from the uncorrected pairwise distances between
234
aligned 16S rRNA gene sequences, using the average neighbor clustering
235
(Schloss and Westcott, 2011), considering a cut-off value of 97% similarity.
236
All OTU-based approaches were performed with software package mothur
237
1.34 (Schloss et al., 2009) as well as the taxonomic assignment of the
238
sequences, performed by the Greengenes-based alignment using default
239
parameters. A list of GenBank accession numbers is provided in Supplemental
240
Table 2.
241
242
RESULTS AND DISCUSSIONS
243
Morphology of colored microbial mats and associated microbe-
244
mineral interactions
245
One of the important factors influencing the microbial diversity of
246
subsurface environments is the mineral microniches they develop on (Jones
247
and Bennett, 2014). In order to broaden our understanding of the interactions
248
of Actinobacteria in volcanic caves and their diversity around the world, an
249
extensive SEM study was performed. Colored microbial mats with different
250
morphologies from Azorean, Canadian, Canarian, Hawaiian and New Mexican
251
In review

12
volcanic caves were investigated (Figure 1A). Abundant white and yellow
252
microbial mats were distinctly visible to the naked eye (Figure 1B, C). These
253
colored mats may consist of large, dense expanses of microorganisms with
254
coarse and irregular edges covering extensive areas of volcanic cave walls and
255
ceilings (Figure 1B) or small colonies spread all over the surface (Figure 1C).
256
Some colonies adopted the form of white spots with irregularly radiate pattern
257
(Figure 1C) or yellow, round and isolated spots with a symmetrical character
258
(Figure 1D). They can grow on the rock surfaces or on secondary mineral
259
deposits, such as ooze-like deposits frequently found in these volcanic caves.
260
In general, the microbial mats have finely granular surface (Figure 1D) and act
261
as water condensation points, being covered with water droplets, particularly
262
during the wet seasons (Figure 1E).
263
SEM images revealed the presence of possible actinobacteria-like
264
structures in most of the volcanic caves from all over the world showing a
265
large variety of microbial morphologies and spore surface ornamentation
266
(Figure 2). To confirm this observation, Sanger and pyrosequencing were
267
performed. In general, these microbial mats were formed by a tangled mass of
268
hyphae, spores, filamentous and coccoid cells (Figure 2A-C). Coccoid
269
elements, with a diameter of about 0.5 µm, are frequently found in close
270
heaps, intermingled with filamentous forms (Figure 2B, E). Most of these
271
masses exhibited characteristic arthrospores of Streptomyces or close relatives
272
with hairy (Figure 2C, F), smoothly (Figure 2H), spiny (Figure 2I) surface
273
ornamentations. Spirals at the end of the aerial mycelium were also observed
274
In review

13
(Figure 2J). A notable feature of some of these bacteria is their filamentous
275
growth with true branching, as depicted for instance in Figure 2H. Chains of
276
coccoid cells resembling beads-on-a-string (Figure 2G) were found within
277
both white and yellow mats. Some other microbial structures were difficult to
278
associate to specific genera or species (Figure 8D, E, G, K, L). In addition,
279
large spheres with lumpy surface connected by a network of hairy filaments
280
and EPS (Figure 2K) or CaCO3 spheres (EDS microanalysis) coated with a
281
filamentous network (Figure 2L) were occasionally observed in the colored
282
microbial mats. Average sizes varied between 10 and 15 µm.
283
The microbial mats studied in this work were found to be involved in
284
microbe-mineral interactions as revealed by SEM investigations (Figure 3).
285
Cell-sized etch pits attributed to dissolution of the substrate under attached
286
cells were noticed (Figure 3A-C). Microboring caused by euendolithic growth
287
of coccoid cells was particularly evident on the silicified substrate, leaving
288
imprints of their surface ornamentation on the mineral grains (Figure 3C).
289
These microbial mats may also figure prominently in the deposition of
290
minerals due to the presence of filaments, some of which are coated with
291
minerals (Figure 3D-F). Among them, reticulated filaments similar to those
292
reported by Melim et al. (2008) and Miller et al. (2012a) were found
293
associated with the white microbial mats from the Kula Kai Caverns of the
294
Kipuka Kanohina Cave Preserve (Hawai'i, U.S.A.) and Falda de La Horqueta
295
cave, in La Palma Island, Spain (Figure 3E,F). All these features evidence
296
microbe-mineral interactions and may represent mineralogical signatures of
297
In review

14
life. Both constructive and destructive mineral features in caves have been
298
recognized as biosignatures valuable for the searching of traces of life on Earth
299
and other planets (Banfield et al., 2001; McLoughlin et al., 2007; Hofmann et
300
al., 2008; Northup et al., 2011).
301
The role of Actinobacteria in biomineralization and rock-weathering
302
processes in caves has been discussed in recent years (Cuezva et al., 2012;
303
Porca et al., 2012; Saiz-Jimenez, 2012). Both processes involve destruction
304
and construction of mineral structures. Destructive processes include
305
dissolution, etching or pitting, whereas constructive processes comprise
306
precipitation of secondary minerals, such as calcite, struvite, witherite and
307
birnessite. In terms of weathering of minerals, the major processes promoted
308
by microorganisms are biochemical and biophysical mechanisms of etching,
309
dissolution and boring occurring via mechanical attachment and secretion of
310
exoenzymes or organic acids (Lee et al., 2012). Extensively etched mineral
311
grains such as calcite and Mg-silicate minerals were found associated with
312
actinobacterial morphologies on coralloid-type speleothems from the Ana
313
Heva volcanic lava tube cave in Chile (Miller et al., 2014). In many cases, it is
314
difficult to determine the exact mechanism by which microorganisms induce
315
mineral dissolution, but the pitting of underlying mineral grains, as shown in
316
Figure 3, illustrates that it does occur.
317
On the other hand, microorganisms may directly precipitate minerals as
318
part of their metabolic activity, and they can also indirectly impact mineral
319
formation by altering the chemical microenvironment such as pH or redox
320
In review

15
conditions or providing nucleation sites for precipitation through the production
321
of organic polymers (Benzerara et al., 2011). Numerous biogenic minerals have
322
been reported in subterranean environments (Sanchez-Moral et al., 2003, 2004;
323
Spilde et al., 2005; De los Ríos et al., 2011; Miller et al., 2012b, 2014), and
324
some of them have been associated with actinobacterial communities. Laiz et
325
al. (2003) found that 61% of the Actinobacteria isolated from Altamira Cave
326
(Spain) produced mineral crystals on culture media. In general, culture and field
327
sample biominerals were composed of calcite, aragonite, Mg-calcite or vaterite.
328
Groth et al. (2001) also tested the ability of cave-dwelling bacteria from Grotta
329
dei Cervi (Italy) for producing mineral crystals. These authors reported
330
extensive mineral production among actinobacteria, which induced the
331
precipitation of calcite (e.g. Brachybacterium sp.) or vaterite (e.g. Rhodococcus
332
sp.). Needle-fiber mats were also related to biomineralization processes by
333
actinomycetes (Cañaveras et al., 1999, 2006). Struvite was formed by
334
actinobacteria isolated from tuff in Roman catacombs (Sanchez-Moral et al.,
335
2003), and witherite, a naturally occurring barium carbonate, was produced by
336
species of the genera Agromyces and Streptomyces isolated from tuff (Sanchez-
337
Moral et al., 2004). Calcium carbonate spheres closely related to dense networks
338
of interwoven filaments of actinobacteria were observed within the colored
339
microbial mats from Azorean, Canarian and Hawaiian volcanic caves (Figure
340
2L). Similar spherical particles were previously reported by Cuezva et al. (2012)
341
and Diaz-Herraiz et al. (2013), who proposed vaterite as their mineralogical
342
phase. According to Cuezva et al. (2012) the grey colonies found on Altamira
343
In review

16
cave walls, dominated by Actinobacteria, were able to bioinduce CaCO3
344
precipitation.
345
346
Actinobacterial diversity found in New Mexico (USA), Hawai'i
347
(USA) and Azores islands (Portugal)
348
The SEM study revealed notable microbial diversity. In order to
349
confirm the presence of Actinobacteria in these volcanic caves and further
350
investigate their diversity, three geographically distinct locations, New Mexico
351
(USA), Hawai’i (USA) and Azores islands (Portugal), were chosen for clone
352
library analysis. A total of 1176 Actinobacteria sequences generated by clone
353
libraries were determined to be of high quality and used in this analysis
354
(Supplemental Table 1). These sequences clustered into 164 OTUs across all
355
locations, belonging to seven orders. Actinomycetales (sequences = 76.7%,
356
OTUs = 52.4%), Euzebyales (9.9%, 8.5%) and Acidimicrobiales (9.6%,
357
17.7%) represented the majority of the OTUs (Figure 4A,B, upper panel).
358
Bifidobacteriales (0.8%, 3.0%), Gaiellales (0.9%, 5.5%), Rubrobacterales
359
(0.5%, 3.0%), and candidate 0319-7L1 (0.5%, 3.0%) represented less than 1%
360
of the sequences (Figure 4A,B, lower panel). Sequences that could not be
361
assigned to taxonomic affiliations were labeled as ‘unclassified’ (1.1%, 6.7%).
362
Singletons and doubletons were the most common OTU type over all (116
363
In review

17
singletons, 23 doubletons). Of the doubletons, 14 had two sequences from the
364
same cave, and 20 had sequences from the same location.
365
Five of the OTUs (3.05%) represented 74.1% of the total number of
366
sequences found. The most predominant OTU (OTU 025) belonged to the
367
Pseudonocardiaceae family, with 593 sequences (50.4%) in 59 of the 82
368
samples. The second most common OTU (OTU 089), also a
369
Pseudonocardiaceae, had 98 sequences (8.3%) in 29 samples, but was not
370
found in Hawai'i. Pseudonocardiaceae was the most commonly found
371
sequence and OTU in each location. This finding is consistent with other cave
372
studies, which found Pseudonocardiaceae to comprise 52% of actinobacterial
373
sequences in Carlsbad Cavern (Barton et al., 2007), 30-50% in three Slovenian
374
limestone caves (Porca et al., 2012), and the most abundant OTU in a
375
limestone cave in China (Wu et al., 2015).
376
OTUs belonging to the orders Actinomycetales, Euzebyales,
377
Acidomicrobiales and Bifidobacteriales were shared by at least two of the
378
three locations under study (Figures 5, Supplemental Figure 1A). These
379
ubiquitous OTUs may represent a core subsurface microbiota, a hypothesis
380
that we will test in the future with more extensive sequencing. Furthermore,
381
caves are not homogeneous habitats: they are characterized by zonal
382
environments according to the distance to entrances (Howarth, 1983, 1993;
383
In review

18
Poulson and White, 1969), passage geometry, and microenvironments, which
384
result from several types of reactions, including microbial processes that often
385
involve redox reactions (Barton and Northup, 2007).
386
The number of shared OTUs in the three locations was relatively low;
387
three out of five belonged to Pseudonocardiaceae and two were Euzebyales
388
(Supplemental Figure 1A). Azores and New Mexico shared six other OTUs,
389
four Pseudonocardiaceae, one Euzebyales and one Bifidobacteriales. Both
390
archipelagos shared two Acidomicrobiales, one Pseudonocardiaceae and one
391
unclassified OTU. Chao 1 estimator suggests that even though a more
392
comprehensive sampling is required to provide a more complete assessment of
393
these microbial communities, our sampling effort was probably enough to
394
describe the cosmopolitan OTUs (Supplemental Figure 1B).
395
None of the sequences recovered were classified as Streptomyces,
396
which was odd, given that Streptomyces are present in almost every other
397
environment studied (i.e. soil, marine, etc., Schrempf, 2006), and were found
398
in cultured isolates from the Azores (Riquelme and Dapkevicius personal
399
communication). We believe this anomaly is due to primer bias. Farris and
400
Olson (2007) showed that many Actinobacteria were not amplified in PCR
401
despite being 100% identical to the universal primers used. While this does not
402
conclusively establish that our sequencing missed Streptomyces that are
403
In review

19
present, it is cause for concern. Future sequencing efforts will utilize
404
Actinobacteria-specific primers to test our hypothesis that Streptomyces are
405
being missed and to better characterize the diversity of the Actinobacteria in
406
caves.
407
Euzebyales emerged as the second most abundant order (number of
408
sequences) in New Mexico and Hawai'i; however, Acidomicrobiales had the
409
second most OTUs in New Mexico and Hawai'i, and was second for both
410
sequences and OTUs in the Azores (Figure 6). Euzebyales was recently
411
described and has two known genera (Kurahashi et al., 2010), and highly
412
similar sequences have been identified from numerous environments (sea
413
cucumbers, saline soils and caves) suggesting this order may be widespread in
414
numerous habitats (Cuezva et al., 2012; Ludwig et al., 2012; Ma and Gong,
415
2013; Velikonja et al., 2014). The Acidimicrobiales order was described by
416
Stackebrandt et al., (1997) and comprises members that are obligate
417
acidophiles, oxidize ferrous iron or reduce ferric iron. It has already been
418
described in caves (Macalady et al., 2007; Ortiz et al., 2013; De Mandal et al.,
419
2014), other volcanic environments (Cockell et al., 2013) and Fe-rich
420
environments (Grasby et al., 2013; Sánchez-Andrea et al., 2011).
421
422
Evaluation of diversity coverage and richness of the clone libraries
423
In review

20
The coverage average estimated for the different locations ranged from
424
78-86%. Due to some variation in sampling effort in each case, a re-sampling
425
analysis was performed, randomly selecting the smallest number of sequences
426
across the different groups (139), 1000 times per each sample, to standardize
427
the values. Diversity indices and estimators are summarized in Table 1a. Non-
428
parametric Shannon and Shannon suggested more diverse communities within
429
New Mexico caves compared to Hawai'i and Azores. Simpson diversity
430
indices suggest the highest diversity values for Hawai'i. All indices agree with
431
the less diverse communities being in Azores. The Shannon index gives more
432
weight to the rare species and Simpson to the dominant ones. Considering the
433
Simpson indexes of the three locations, the community composition in Azores
434
caves would include more cosmopolitan species with high abundance and
435
Hawai'i caves would be composed of phylotypes with narrower distribution. In
436
islands, population size and genetic diversity tend to be limited due to the
437
smaller extension of the habitats. Comparable taxa–area relationships (Bell et
438
al., 2005) and distance–decay relationships for microbes and larger organisms
439
were found to be significant although with variations in the rates of the
440
processes (reviewed by Soininen, 2012; Green and Bohannan, 2006).
441
However, we found differences between the diversity indices for Azores and
442
Hawai'i, what could be related to differences in island size, isolation and age
443
In review

21
of lava flows. We should be aware that the amount of data available is still
444
small and that further studies may still reveal different trends.
445
446
Phylogenetic analysis
447
When the representative sequences from each OTU were compared to
448
known sequences in GenBank, 17 out of 164 OTUs (10%) shared ≤90%
449
identity with known sequences in GenBank (Figure 7). Fifty two percent of
450
the OTUs shared between 91% and 96% identity and 38% shared over 97%
451
identity with known sequences. The most novel OTUs were mostly singletons,
452
and were classified as Pseudonocardiaceae (four OTUs), Rubrobacteraceae
453
(one OTU), Bifidobacteriales (five OTUs) and unclassified (seven OTUs).
454
They were found in all four locations, however, more of the OTUs were found
455
in the Azorean islands (13 out of 100) than in Hawai'i (2 out of 30) or New
456
Mexico (3 out of 54). Physical isolation is an important driver of microbial
457
evolution (Papke and Ward, 2004); thus, island isolation would promote
458
unique evolutionary forces that result in the development of a novel genetic
459
reservoir. However, in our results we did not observe significant differences
460
between continental and island territories according to genetic novelty.
461
An approximate maximum likelihood tree shows the relationship
462
between the sequences and occurrence of OTUs (Figure 5). For this analysis
463
In review

22
Pico and Terceira were considered separate locations. Gaiellaceae-like
464
sequences were found in three New Mexico and the Azores, but not Hawai'i.
465
All but one of the sequences were singletons. Gaiellaceae, another recently
466
described family, was originally found in a water borehole, and sequences
467
from this family have subsequently been found in soil, volcanic soil, thermal
468
springs and marine ascidians (Albuquerque et al., 2011; Kim et al., 2014;
469
Rozanov et al., 2014; Steinert et al., 2015). Rubrobacterales occurred in the
470
New Mexico and Hawai'i samples. The order Actinomycetales has many
471
polytomies with most of them occurring in the samples from Hawai'i, Pico,
472
and Terceira. These samples are either unresolved parts of the tree due to
473
missing data or represent rapid speciation in the Actinomycetales.
474
Representatives of Euzebyales were found in all four locations (Figure 5). The
475
different clades suggest there is significant diversity within the sequences
476
found.
477
While we acknowledge the limitation of our study to capture the full
478
range of diversity in these sites, the high number of singletons found in this
479
study suggests that there are Actinobacteria belonging to the rare biosphere in
480
caves. The rare biosphere has been shown to influence both alpha and beta
481
diversity, exhibiting unique geographic patterns (Lynch and Neufeld, 2015).
482
In review

23
With over two thirds of our OTUs being singletons and most of the
483
doubletons from one location, there is evidence to suggest endemism in cave
484
Actinobacteria. Endemism in caves has been documented for obligate cave
485
fauna in the United States and the Azores (Culver et al., 2003; Reboleira et al.,
486
2012). Furthermore, studies of Actinobacteria in other environments have been
487
shown to display endemism (Wawrik et al., 2007; Valverde et al., 2012). The
488
combination of rare and endemic Actinobacteria, together with their
489
abundance in caves, support the idea that caves are a good location to further
490
test hypotheses regarding bacterial biogeography as well as to look for novel
491
actinobacterial metabolites. Rigorous testing will require that future studies be
492
conducted with next generation sequencing to comprehensively sample the
493
diversity present in these habitats.
494
495
16S rRNA gene amplicon library preparation, sequencing, bioinformatics
496
and OTU-based analysis in Azorean volcanic caves
497
The observed structure of the microbial communities in volcanic caves
498
in the three locations is consistent with bacterial communities composed of
499
consortia of few cosmopolitan members and a high number of low abundant
500
phylotypes. To test whether this structure could be biased by the fact of having
501
In review

24
a limited number of sequences, a pyrosequencing approach was performed
502
with the same sample points considered for clone libraries in Azores.
503
Actinobacterial sequences amplified using the universal primers were
504
identified and after quality control and filtering of the crude pyrotags, 19,476
505
sequences with good quality were retained, consisting of 906 unique
506
sequences. The average sequence length was 247.5 bp (range 233-275; median
507
247.1; sd 4.1). After clustering, a total of 382 OTUs were obtained.
508
Nine orders were found in Azorean caves with pyrosequencing, the
509
seven previously found, i.e. candidate 0319-7L1 (sequences = 0.4%, OTU =
510
2.9%), Acidimicrobiales (1.2%, 1.6%), Actinomycetales (92.6%, 62.8%),
511
Bifidobacteriales (0.7% , 4.5%), Euzebyales (2.7%, 4.5%), Gaiellales (1.1%,
512
8.4%), Rubrobacterales (0.04%, 0.3%), plus Coriobacteriales (0,3%, 3.4%)
513
and Solirubrobacterales (0.3%, 4.5%) (Figure 8). While Rubrobacterales was
514
found in the clone libraries, it was only found in New Mexico and Hawai'i
515
(Figure 5). Amplicon sequencing revealed this order to be present in the
516
Azores as well, highlighting the importance of pyrosequencing to capture the
517
full range of diversity in these samples. Actinomycetales and Gaiellales orders
518
showed an increase in the percentage of sequences and OTUs recovered;
519
Bifidobacteriales had a higher percentage of OTUs. All other orders displayed
520
In review

25
lower percentages both for sequences and OTUs. Unclassified sequences
521
represented 0.7% and 7.3%, respectively.
522
The amplicon libraries approach showed a more complete picture of
523
the subterranean diversity in Azorean volcanic caves. Rubrobacterales
524
comprised a group of novel OTUs, with all sequences sharing no more than
525
92% similarity with known sequences in GenBank, as well as
526
Solirubrobacterales, with all of the sequences ranging between 90-95%
527
similarity (Stackebrandt et al., 1997; Reddy and Garcia-Pichel, 2009).
528
Rubrobacterales was first described in cave environments in Niu Cave (Zhou
529
et al., 2007), and were also recovered from speleothems in Kartchner Caverns.
530
This order includes members with heat, cold, dryness and high radiation
531
resistance, found in high number in biodeteriorated monuments (Gurtner et al.,
532
2000; Jurado et al., 2012) and volcanic environments (Cockell et al., 2013).
533
Solirubrobacterales have also been described in caves (Paterson, 2012; De
534
Mandal, 2014) and in other volcanic environments (Gomez-Alvarez et al.,
535
2007; Cockell et al., 2013). Coriobacteriales (Stackebrandt et al., 1997; Gupta
536
et al., 2013) showed a high percentage of sequences, 89.1%, with more than
537
97% similarity. This order was previously described in cave habitats in
538
speleothem formations in Kartchner Caverns (Ortiz et al., 2013), and in Lower
539
Kane cave (Paterson, 2012).
540
In review

26
541
Evaluation of diversity coverage and richness of the amplicon libraries
542
Sampling completeness assessed by Good's coverage estimator for
543
each data set returned values above 98% (Table 1b). Diversity indices revealed
544
a higher diversity at Pico Island compared to Terceira Island as well as chao
545
richness estimator (Table 1b).
546
The dominance of the Pseudonocardiaceae family compared to any
547
other member of the microbial community is remarkable, in accordance with
548
results from both clone and amplicon libraries. Pseudonocardiaceae
549
encompases a wide array of rare Actinomycetes, many of which can produce
550
secondary metabolites (Tiwari and Gupta, 2013). While we acknowledge that
551
this finding may be in part the result of primer bias, the prevalence of this
552
family is not uncommon in caves (Barton et al., 2007, Porca et al., 2012; Wu
553
et al., 2015). Little is known of role these bacteria play in most ecosystems,
554
however the family encompases a wide variety of metabolic pathways and
555
physiologies (Huang and Goodfellow, 2011). Most of our sequences were
556
unable to be classified at the genus level, leaving some doubt as to the true role
557
of this group of bacteria in volcanic caves. However, the ubiquity of this
558
family in cave studies emphasizes the need for further molecular studies with
559
improved primers to capture Actinobacteria diversity and cultivation of
560
In review

27
members of this family found in subterranean bacterial biofilms. An
561
examination of the communities in situ combined with metatranscriptome
562
analysis would shed light on the question of this group's role in volcanic cave
563
ecosystems.
564
565
566
CONCLUSIONS
567
568
Our collective attempt to better understand actinobacterial diversity
569
and functions in volcanic caves led us to observe patterns of diversity and
570
novelness through a range of data obtained from 454 pyrosequencing to
571
cloning. To date, within the realm of actinobacterial community study, our
572
work is one of the largest sampling efforts in volcanic caves from different
573
parts of the world including Spain, Portugal, USA and Canada. The
574
sequencing effort, both in clone and amplicon libraries, represents one of the
575
most comprehensive studies of Actinobacteria in volcanic caves around the
576
world. The clone libraries illustrate the novelness and phylogenetic
577
relationship of Actinobacteria in volcanic caves from three geographically
578
distant locations. The amplicon libraries of the Azorean sequences gave a
579
more in-depth view of the Actinobacteria communities and revealed more
580
diversity than has previously been described. Both methods showed large
581
numbers of newly described orders, and a dominance of Actinomycetales.
582
Together they provide an outline of the community structure of Actinobacteria
583
In review

28
in caves, and highlight the importance of caves as a source of rare and novel
584
Actinobacteria.
585
Through scanning electron microscopy examinations, we learned about
586
bacterial morphology, their relationships and possible contribution of the
587
Actinobacteria to cave environment. The identification of Ca-rich elements
588
coated within some of the filamentous networks in the colored microbial mats
589
suggests a possible role of Actinobacteria in calcium deposition. Both
590
constructive and destructive mineral features, such as biominerals, cell
591
imprints, microboring and mineralized filaments were some of the
592
biosignatures found associated with samples studied herein. We can thus
593
consider that volcanic caves on Earth are plausible repositories of terrestrial
594
biosignatures where we can look for evidence of early life.
595
Beyond contributing to understanding cave microbial ecology,
596
community and microbial roles and related function in such extreme
597
subsurface habitats, our study hopes to initiate more study in such an
598
interesting and understudied frontier of the Earth, where unique compounds
599
could be isolated and used as important sources of industrial processes.
600
601
ACKNOWLEDGEMENTS
602
The authors acknowledge the Spanish Ministry of Economy and
603
Competitiveness (project CGL2013-41674-P) and FEDER Funds for financial
604
support. A.Z. Miller acknowledges the support from the Marie Curie Intra-
605
European Fellowship of the European Commission’s 7th Framework
606
In review

29
Programme (PIEF-GA-2012-328689). C. Riquelme was funded by the
607
Regional Fund for Science and Technology and Pro-Emprego program of the
608
Regional Government of the Azores, Portugal [M3.1.7/F/013/2011,
609
M3.1.7/F/030/2011]. Her work was partly supported by National funds from
610
the Foundation for Science and Technology of the Portuguese Government,
611
[Understanding Underground Biodiversity: Studies in Azorean Lava Tubes
612
(reference PTDC/AMB/70801/2006]. The authors would like to thank the
613
TRU Innovation in Research Grant, TRU UREAP Fund, Western Economic
614
Diversification Canada Fund, Kent Watson (assisted with the Helmcken Falls
615
Cave sample collection), Derrick Horne (UBC BioImaging Facility for the
616
SEM work). We acknowledged the Canadian Ministry of Forests, Lands and
617
Natural Resource Operations for Park Use Permit#102172. This work was also
618
supported by the Cave Conservancy of the Virginias, the Graduate Research
619
Allocation Committee at UNM Biology, UNM Biology Grove Scholarship,
620
the Student Research Allocation Committee at UNM, the National
621
Speleological Society, the New Mexico Space Grant Consortium, the New
622
Mexico Alliance for Minority Participation Program, the New Mexico
623
Geological Society, and Kenneth Ingham Consulting. We acknowledge
624
support from the UNM Molecular Biology Facility, which is supported by
625
NIH grant number P20GM103452. The authors also wish to thank Fernando
626
Pereira, Ana Rita Varela, Pedro Correia, Berta Borges and Guida Pires for
627
help during field and lab work in the Azores. The authors gratefully
628
acknowledge the photographic contributions of Kenneth Ingham and Pedro
629
In review

30
Cardoso and Michael Spilde (SEM images). The authors would like to thank
630
Dr. Steven Van Wagoner (TRU) and Drs. Julian Davies and Vivian Miao
631
(UBC) for their invaluable comments in manuscript preparation.
632
References:
633
634
Albuquerque L., Franc¸ L. Rainey, F. A., Schumann, P., Nobre, M. F. and da
635
Costa M.. S. (2011). Gaiella occulta gen. nov., sp. nov., a novel
636
representative of a deep branching phylogenetic lineage within the class
637
Actinobacteria and proposal of Gaiellaceae fam. nov. and Gaiellales ord.
638
nov. Syst. Appl. Microbiol. 39, 595– 599.
639
Altschul, S. F., Madden, T.L., Schäffer, A. A., Zhang, J., Zhang, Z., et al.
640
(1997). Gapped BLAST and PSI-BLAST: a new generation of protein
641
database search programs. Nucleic Acids Res. 25, 3389-3402.
642
Banfield J. F., Moreau J. W., Chan C. S., Welch, S. A., and Little, B. (2001).
643
Mineralogical biosignatures and the search for life on Mars. Astrobiology.
644
1, 447-465.
645
Barton H. A., and Northup D. E. (2007). Geomicrobiology in cave
646
environments: past, current and future perspectives. J. Cave Karst Stud. 69,
647
163–178.
648
Barton H. A., Taylor N. M., Kreate M. P., Springer A. C., Oehrle S. A., and
649
Bertog J. L. (2007). The impact of host rock geochemistry on bacterial
650
community structure in oligotrophic cave environments. Int. J. Speleol.
651
36,:93-104.
652
Barton H. A., Giarrizzo J. G., Suarez P., Robertson C. E., Broering M. J.,
653
Banks, E. D. et al. (2014). Microbial diversity in a Venezuelan
654
orthoquartzite cave is dominated by the Chloroflexi (Class
655
Ktedonobacterales) and Thaumarchaeota Group I.1c. Front. Microbiol. 5,
656
00615 DOI: 10.3389/fmicb.2014.00615.
657
Benzerara, K., Miot, J., Morin, G., Ona-Nguema, G., Skouri-Panet, F., and
658
Ferard, C. (2011). Significance, mechanisms and environmental
659
implications of microbial biomineralization. C. R. Geosci. 343, 160-167.
660
Bell T., Ager D., Song J. I., Newman J. A., Thompson I. P., et al. (2005).
661
Larger islands house more bacterial taxa. Science. 308, 1884-1884.
662
Boston, P. J., Spilde, M.N., Northup, D.E., Melim, L.A., Soroka, D.S., Kleina,
663
L.G. et al. (2001). Cave biosignature suites: microbes, minerals, and Mars.
664
Astrobiology 1, 25-55.
665
Brantner J. S., Haake Z. J., Burwick J. E., Menge C. M., Hotchkiss S. T., and
666
Senko J. (2014). Depth-dependent geochemical and microbiological
667
gradients in Fe(III) deposits resulting from coal mine-derived acid mine
668
drainage. Front. Microbiol. 5, (00215).
669
Cañaveras, J. C., Hoyos, M., Sanchez-Moral, S., Sanz-Rubio, E., Bedoya, J.,
670
Soler, V. et al. (1999). Microbial communities associated to
671
In review

31
hydromagnesite and needle fiber aragonite deposits in a karstic cave
672
(Altamira, Northern Spain). Geomicrobiol. J. 16, 9-25.
673
Cañaveras, J. C., Cuezva, S., Sanchez-Moral, S., Lario, J., Laiz, L., Gonzalez,
674
J. M., et al. (2006). On the origin of fiber calcite crystals in moonmilk
675
deposits. Naturwissenschaften 3, 27-32.
676
Caporaso J.G., Kuczynski J., Stombaugh J., Bittinger K., Bushman F. D.,
677
Costello E. K., et al. (2010). QIIME allows analysis of high-throughput
678
community sequencing data. Nat. Methods. 7,335-336.
679
doi:10.1038/nmeth.f.303.
680
Chao A. (1984). Non-parametric estimation of the number of classes in a
681
population. Scand. J. Stat. 11, 265–270.
682
Chao A., and Shen T. J. (2003). Nonparametric estimation of Shannon’s index
683
of diversity when there are unseen species in sample. Environ. Ecol. Stat.
684
10, 429–443.
685
Cheeptham, N., Sadoway, T., Rule, D., Watson, K., Moote, P., Soliman, L., et
686
al. (2013). Cure from the cave: volcanic cave actinomycetes and their
687
potential in drug discovery. Int. J. Speleol. 42, 35-47.
688
doi:http://dx.doi.org/10.5038/1827-806X.42.1.5.
689
Cockell, C.S., Kelly, L.C. and Marteinsson, V. (2013). Actinobacteria-An
690
Ancient Phylum Active in Volcanic Rock Weathering. Geomicrobiol. J.
691
30:8, 706-720, DOI: 10.1080/01490451.2012.758196.
692
Cuezva, S., Fernandez-Cortes, A., Porca, E., Pašić, L., Jurado, V., Hernandez-
693
Marine, M., et al. (2012). The biogeochemical role of Actinobacteria in
694
Altamira Cave, Spain. FEMS Microbiol. Ecol. 81, 281–290.
695
Culver D. C., Christman M. C., Elliott W. R., Hobbs H. H., and Reddell J. R.
696
(2003). The North American obligate cave fauna: Regional patterns.
697
Biodivers. Conserv. 12, 441-468.
698
De los Ríos, A., Bustillo, M. A., Ascaso, C., and Carvalho, M.R. (2011).
699
Bioconstructions in ochreous speleothems from lava tubes on Terceira
700
Island (Azores). Sediment. Geol. 236, 117-128.
701
De Mandal S, Sanga Z, and Nachimuthu S. K. (2014). Metagenomic analysis
702
of bacterial community composition among the cave sediments of Indo-
703
Burman biodiversity hotspot region. Peer J. PrePrints 2 : e631v1.
704
doi.org/10.7287/peerj.preprints.631v1.
705
Diaz-Herraiz, M., Jurado, V., Cuezva, S., Laiz, L., Pallecchi, P., Tiano, P. et
706
al. (2013). The actinobacterial colonization of Etruscan paintings. Sci. Rep.
707
3, 1440.
708
Dowd, S. E., Y. Sun, Wolcott R. D., Domingo A., and Carroll, J. A. (2008).
709
Bacterial tag-encoded FLX amplicon pyrosequencing (bTEFAP) for
710
microbiome studies: bacterial diversity in the ileum of newly weaned
711
Salmonella-infected pigs. Foodborne Pathog. Dis. 5, 459-472.
712
Edgar, R.C. (2010). Search and clustering orders of magnitude faster than
713
BLAST. Bioinformatics. 26, 2460-2461. doi:
714
10.1093/bioinformatics/btq461
715
In review

32
Farris, M.H. and Olson, J.B. (2007). Detection of Actinobacteria cultivated
716
from environmental samples reveals bias in universal primers. Lett. Appl.
717
Microbiol. 45, 376-381.
718
Gomez-Alvarez, V., King, G.M., and Nüsslein, K. (2007). Comparative
719
bacterial diversity in recent Hawaiian volcanic deposits of different ages.
720
FEMS Microbiol. Ecol. 60, 60–73.
721
Good, I.J. (1953). The population frequencies of species and the estimation of
722
population parameters. Biometrika 40, 237–264.
723
Goodfellow, M., and Fiedler, H.P. (2010). A guide to successful
724
bioprospecting: informed by actinobacterial systematic. Anton Leeuw. 98,
725
119–142 doi: 10.1007/s10482-010-9460-2.
726
Grasby, S.E., Richards, B.C., Sharp, C.E., Brady, A.L., Jones, G.M. and
727
Dunfield, P.F. (2013). The Paint Pots, Kootenay National Park, Canada —
728
a natural acid spring analogue for Mars. Can. J. Earth Sci. 50: 94–108.
729
Green, J. and Bohannan, B.J.M. (2006). Spatial scaling of microbial
730
biodiversity. Trends Ecol. Evol. 21: 501–507.
731
Groth I., Schumann P., Schuetze B., Augsten K., Kramer I., and Stackebrandt
732
E. (1999). Beutenbergia cavernae gen. nov., sp. nov., an L-lysine-
733
containing actinomycete isolated from a cave. Int. J. Syst. Evol. Microbiol.
734
49, 1733-1740.
735
Groth, I., Schumann, P., Laiz, L., Sanchez-Moral, S., Cañaveras, J.C., and
736
Saiz-Jimenez, C. (2001). Geomicrobiological study of the Grotta dei Cervi,
737
Porto Badisco, Italy. Geomicrobiol. J. 18, 241-258.
738
Guo, X., Liu, N., Li, X., Ding, Y., Shang, F., Gao, Y., et al. (2015). Red soils
739
harbor diverse culturable actinomycetes that are promising sources of novel
740
secondary metabolites. Appl. Environ. Microbiol. 81, 3086 –3103.
741
doi:10.1128/AEM.03859-14.
742
Gupta, R.S., Chen, W.J., Adeolu, M., and Chai, Y. (2013). Molecular
743
signatures for the class Coriobacteriia and its different clades; proposal for
744
division of the class Coriobacteriia into the emended order
745
Coriobacteriales, containing the emended family Coriobacteriaceae and
746
Atopobiaceae fam. nov., and Eggerthellales ord. nov., containing the family
747
Eggerthellaceae fam. nov. Int. J. Syst. Evol. Microbiol. 63, 3379-3397.
748
Gurtner, C., Heyrman, J., Piñar, G., Lubitz, W., Swings, J., and Rölleke S.
749
(2000). Comparative analyses of the bacterial diversity on two different
750
biodeteriorated wall paintings by DGGE and 16S rDNA sequence analysis.
751
Int. Biodeter. Biodegr. 46, 229–239.
752
Hathaway, J.J.M., Garcia, M.G., Moya, M., Spilde, M.N., Stone, F. D.,
753
Dapkevicius, M. D. L. N. E., et al. (2014). Comparison of Bacterial
754
Diversity in Azorean and Hawaiian Lava Cave Microbial Mats.
755
Geomicrobiol. J. 31, 205-220.
756
Hofmann, B.A., Farmer, J.D., von Blanckenburg, F., and Fallick, A.E.
757
(2008). Subsurface filamentous fabrics: an evolution of origins based on
758
morphological and geochemical criteria, with implications for
759
exopaleontology. Astrobiology. 8, 87-117.
760
In review

33
Howarth, F.G. (1983). Ecology of cave arthropods. Ann. Rev. Entomol. 28,
761
365-389.
762
Howarth, F.G. (1993). High-stress subterranean habitats and evolutionary
763
change in cave-inhabiting arthropods. Am. Nat. 142, S65-S77.
764
Huang, Y. and Goodfellow, M. (2011) “Genus I. Pseudonocardia Henson
765
1957, 408AL emeds. Park, Park, Lee and Kim 2008, 2477” in Bergey’s
766
Manual of Systematic Bacteriology. ed Whitman W., Goodfellow M.,
767
Kämpfer P., Busse H.-J., Trujillo, M. et al. (New York: Springer) 5, 1305-
768
1323.
769
Jones, A.A., and Bennett P.C. (2014) Mineral Microniches Control the Diversity
770
of Subsurface Microbial Populations. Geomicrobiol. J. 31, 246–261.
771
Jurado, V. Laiz L., Gonzalez J. M., Hernandez-Marine M., Valens M. and
772
Saiz-Jimenez C. (2005a). Phyllobacterium catacumbae sp. nov., a member
773
of the order ‘Rhizobiales’ isolated from Roman Catacombs. Int. J. Syst.
774
Evol. Microbiol. 55, 1487–1490.
775
Jurado, V., Groth, I., Gonzalez, J.M., Laiz, L., Schuetze, B., and Saiz-Jimenez
776
C. (2005b). Agromyces subbeticus sp. nov., isolated from a cave in southern
777
Spain. Int. J. Syst. Evol. Microbiol. 55, 1897–1901.
778
Jurado, V., Miller, A.Z., Alias-Villegas, C., Laiz, L., and Saiz-Jimenez C
779
(2012). Rubrobacter bracarensis sp. nov., a novel member of the genus
780
Rubrobacter isolated from a biodeteriorated monument. Syst, Appl.
781
Microbiol. 35 (5), 306-9.
782
Kim Y., Yoon H., Kim M., Nam Y., Kim H., Seo Y., et al. (2014).
783
Metagenomic analysis of bacterial communities on Dokdo Island. J. Gen.
784
Appl. Microbiol. 60, 65-74.
785
Kurahashi, M., Fukunaga, Y., Sakiyama, Y., Harayama, S., and Yokota, A.
786
(2010). Euzebya tangerina gen. nov., sp. nov., a deeply branching marine
787
actinobacterium isolated from the sea cucumber Holothuria edulis, and
788
proposal of Euzebyaceae fam. nov., Euzebyales ord. nov. and
789
Nitriliruptoridae subclassis nov. Int. J. Syst. Evol. Microbiol. 60, 2314–
790
2319.
791
Laiz, L., Gonzalez, J.M., and Saiz-Jimenez, C. (2003). Microbial communities
792
in caves: Ecology, physiology, and effects on paleolithic paintings. In:
793
Koestler, R.J., Koestler, V.R., Carola, A.E., Nieto-Fernández, F.E. (Eds)
794
Art, Biology, and Conservation: Biodeterioration of works of Art. The
795
Metropolitan Museum of Art, New York, pp. 210-225.
796
Lee, S.D., Kang, S., and Hah, Y.C. (2000). Hongia gen. nov., a new genus of
797
the order Actinomycetales. Int. J. Syst. Evol. Microbiol. 50, 191–199.
798
Lee, S.D., Kim, E.S., Min, K.L., Lee, W.Y., Kang, S.O., and Hah, Y.C.
799
(2001). Pseudonocardia kongjuensis sp. nov., isolated from a gold mine
800
cave. Int. J. Syst. Evol. Microbiol. 51, 1505–1510.
801
Lee, S. D. (2006). Actinocorallia cavernae sp. nov., isolated from a natural
802
cave in Jeju, Korea. Int. J. Syst. Evol. Microbiol. 56, 1085–1088.
803
Lee, N.M., Meisinger, D.B., Aubrecht, R., Kovacik, L., Saiz-Jimenez, C.,
804
Baskar, S., et al. (2012). Caves and karst environments. in: Life at
805
In review

34
Extremes: Environments, Organisms and Strategies for Survival (ed. E.M.
806
Bell), CAB Internacional, pp. 320-344.
807
Léveillé, R.J., and Datta, S. (2010). Lava tubes and basaltic caves as
808
astrobiological targets on Earth and Mars: a review. Planet Space Sci. 58,
809
592-598.
810
Ludwig, W., Euzeby, J., Schumann, P., Busse, H.J., Trujillo, M.E., et al.
811
(2012) “Road map of the phylum Actinobacteria” in Bergey’s Manual of
812
Systematic Bacteriology. ed Whitman W., Goodfellow M., Kämpfer P.,
813
Busse H.-J., Trujillo, M. et al. (New York: Springer) 5, 1-28.
814
Lynch, M.D., and Neufeld, J.D. (2015) Ecology and exploration of the rare
815
biosphere. Nat. Rev. Microbiol. 13, 217-229.
816
Ma, B., and Gong, J. (2013) A meta-analysis of the publicly available bacterial
817
and archaeal sequence diversity in saline soils. World J. Microbiol.
818
Biotechnol. 29, 2325–2334.
819
Macalady, J.L., Jones, D.S., and, Lyon, E.H. (2007). Extremely acidic,
820
pendulous cave wall biofilms from the Frasassi cave system, Italy. Environ
821
Microbiol. 9,1402-1414.
822
Maidak, B.L., Cole, J.R., Lilburn, T.G., Parker Jr., C.T., Saxman, P.R., Farris,
823
R.J., Olsen, G.J., Schmidt, T.M., and Tiedje, J.M. 2001. The RDP-II
824
(Ribosomal Database Project). Nucl. Acids Res. 29,173–174.
825
McDonald, D, Price M. N., Goodrich J., Nawrocki E. P ., DeSantis T. Z.,
826
Probst A., et al.(2012). An improved greengenes taxonomy with explicit
827
ranks for ecological and evolutionary analyses of bacteria and archaea.
828
ISME J. 6, 610–18.
829
McLoughlin, N., Brasier, M.D., Wacey, D., Green, O.R., and Perry, R.S.
830
(2007). On biogenicity criteria for endolithic microborings on early Earth
831
and beyond. Astrobiology. 7, 10-26.
832
McMurdie, P. J. and Holmes, S. (2013). phyloseq: An R Package for
833
Reproducible Interactive Analysis and Graphics of Microbiome Census
834
Data. PloS one 8, e61217. doi: 10.1371/journal.pone.0061217.
835
Melim, L.A., Northup, D.E., Spilde, M.N., Jones, B., Boston, P.J., and Bixby,
836
R.J. (2008). Reticulated filaments in cave pool speleothems: microbe or
837
mineral? J. Cave Karst Stud. 70, 135-141.
838
Miao, V., and Davies, J. (2010). Actinobacteria: the good, the bad, and the
839
ugly. Anton Leeuw. 98, 143-150. doi: 10.1007/s10482-010-9440-6
840
Miller, A.Z., Hernández-Mariné, M., Jurado, V., Dionísio, A., Barquinha, P.,
841
Fortunato, E. et al. (2012a). Enigmatic reticulated filaments in subsurface
842
granite. Environ. Microbiol. Rep. 4, 596-603.
843
Miller, A.Z., Dionísio, A., Sequeira Braga, M.A., Hernández-Mariné, M.,
844
Afonso, M.J., Muralha, V.S.F. et al. (2012b). Biogenic Mn oxide minerals
845
coating in a subsurface granite environment. Chem. Geol. 322-323, 181-
846
191.
847
Miller A. Z., Pereira, M. F. C., Forti, P., Dionisio, A., and Saiz-Jimenez, C.
848
(2014). Siliceous speleothems and associated microbe-mineral interactions
849
from Ana Heva Lava tube in Easter Island (Chile). Geomicrobiol. J. 31,
850
236-245.
851
In review

35
Moya, M., Garcia M. G., Spilde M.N., and Northup D. E. (2009) Composition
852
of Bacterial mats in El Malpais, National Monument, New Mexico, USA:
853
comparison and contrasts with Bacterial communities in Hawai'i lava tubes.
854
Proc. Int. Congr. Speleol. 2, 709-713.
855
Nawrocki, E.P., Kolbe, D.L., and Eddy, S.R. (2009). Infernal 1.0: inference of
856
RNA alignments. Bioinformatics 25, 1335–1337.
857
Niyomyong, N. Pathom-aree W., Arinthip Thamchaipenet A., and Duangmal
858
K. (2012) Actinomycetes from Tropical Limestone Caves. Chiang Mai J.
859
Sci. 39, 373-388.
860
Northup, D.E., Snider, J.R., Spilde, M.N., Porter, M.L., Van de Kamp, J.L.,
861
Boston, P.J., Nyberg, A.M., and Bargar, J.R.. 2010. Diversity of rock
862
varnish Bacterial communities from Black Canyon, New, Mexico. J.
863
Geophys. Res. 115,G02007. DOI: 10.1029/2009JG001107
864
Northup, D.E., Melim, L.A., Spilde, M.N., Hathaway, J.J.M., Garcia, M.G.,
865
Moya, M. et al. (2011). Lava cave microbial communities within mats and
866
secondary mineral deposits: Implications for life detection on other planets.
867
Astrobiology 11, 1-18.
868
Ortiz, M., Neilson, J.W., Nelson, W. M., Legatzki, A., Byrne, A., Yu, Y., et
869
al. (2013). Profiling Bacterial Diversity and Taxonomic Composition on
870
Speleothem Surfaces in Kartchner Caverns, AZ. Microb. Ecol. 65, 371–383
871
Papke RT, and Ward, DM. (2004) The Importance of Physical Isolation to
872
Microbial Diversity. FEMS Microbiol. Ecol. 48, 293-303.
873
Pašić, L., Kovče, B., Sket, B., and Herzog-Velikonja, B. (2010). Diversity of
874
microbial communities colonizing the walls of a Karstic cave in Slovenia.
875
FEMS Microbiol. Ecol. 71, 50-60.
876
Paterson, A.T. (2012). The Distribution and Diversity of Functional Gene
877
Pathways Controlling Sulfur Speciation in Lower Kane Cave, Wyoming.
878
Master's Thesis. Department of Geology and Geophysics. Louisiana State
879
University and Agricultural and Mechanical College.
880
Porca E., Jurado V., Žgur-Bertok D., Saiz-Jimenez C., and Pašić, L. (2012).
881
Comparative analysis of yellow microbial communities growing on the
882
walls of geographically distinct caves indicates a common core of
883
microorganisms involved in their formation. FEMS Microbiol. Ecol. 81,
884
255–266.
885
Poulson, T.L., and White, W.B. (1969). The cave environment. Science.
886
165(3897), 971-981.
887
Price, M.N., Dehal, P.S., and Arkin, A.P. (2009). FastTree: Computing Large
888
Minimum-Evolution Trees with Profiles instead of a Distance Matrix. Mol.
889
Biol. Evol. 26:1641-1650, doi:10.1093/molbev/msp077.
890
Price, M.N., Dehal, P.S., and Arkin, A.P. (2010). FastTree 2 -- Approximately
891
Maximum-Likelihood Trees for Large Alignments. PLoS ONE. 5, e9490.
892
doi:10.1371/journal.pone.0009490.
893
Provencio, P.P., and Polyak, V.J. (2010). Iron Oxide-Rich Filaments: Possible
894
Fossil Bacteria in Lechuguilla Cave, New Mexico. Geomicrobiol. J. 18,
895
297-309.
896
In review

36
Quintana, E.T., Badillo, R.F., and Maldonado, L. A. (2013). Characterisation
897
of the first actinobacterial group isolated from a Mexican extremophile
898
environment. Anton Leeuw. 104, 63–70.
899
R Core Team (2015). R: A language and environment for statistical
900
computing.Vienna, Austria: R Foundation for Statistical Computing.
901
Available online at: http://www.R-project.org/.
902
Reboleira, A.S.P.S., Borges, P.A.V., Gonçalves, F., Serrano, A.R.M., and
903
Oromí, P. (2012). The subterranean fauna of a biodiversity hotspot region -
904
Portugal: an overview and its conservation. Int. J. Speleol. 40, 23-37.
905
Reddy, G.S.N., and Garcia-Pichel, F. (2009). Description of Patulibacter
906
americanus sp. nov., isolated from biological soil crusts, emended
907
description of the genus Patulibacter Takahashi et al. 2006 and proposal of
908
Solirubrobacterales ord. nov. and Thermoleophilales ord. nov. Int. J. Syst.
909
Evol. Microbiol. 59:87-94.
910
Rozanov, A.S., Bryanskaya, A.V., Malup, T.K., Meshcheryakova, I.A.,
911
Lazareva, E.V., Taran, O.P., et al. (2014). Molecular analysis of the benthos
912
microbial community in Zavarzin thermal spring (Uzon Caldera,
913
Kamchatka, Russia). BMC Genomics. 15, S12.
914
Saiz-Jimenez, C. (2012). Microbiological and environmental issues in show
915
caves. World J. Microbiol. Biotechnol. 28, 2453-2464.
916
Sánchez-Andrea, I., N. Rodriguez, R. Amils, and J. L. Sanz. (2011). Microbial
917
Diversity in Anaerobic Sediments at Rio Tinto, a Naturally Acidic
918
Environment with a High Heavy Metal Content. Appl. Environ. Microbiol.
919
77, 6085-6093.
920
Sanchez-Moral, S., Bedoya, J., Luque, L., Cañaveras, J.C., Jurado, V., Laiz, L.
921
et al. (2003). Biomineralization of different crystalline phases by bacteria
922
isolated from catacombs. In: Saiz-Jimenez, C. (Ed.), Molecular Biology and
923
Cultural Heritage, Balkema, Lisse, pp. 179-185
924
Sanchez-Moral, S., Luque, L., Cañaveras, J.C., Laiz, L., Jurado, V., and Saiz-
925
Jimenez, C. (2004). Bioinduced barium precipitation in St Callixtus and
926
Domitilla catacombs. Ann. Microbiol. 54, 1-12.
927
Schloss, P.D., Westcott, S.L., Ryabin, T., Hall, J.R., Hartmann, M., Hollister,
928
E.B., et al. (2009). Introducing mothur: Open-source, platform-
929
independent, community-supported software for describing and comparing
930
microbial communities. Appl. Environ. Microbiol. 75, 7537–7541.
931
Schloss, P.D., Gevers, D., and Westcott, S.L., (2011). Reducing the Effects of
932
PCR Amplification and Sequencing Artifacts on 16S rRNA-Based Studies.
933
PLoS ONE 6, e27310. doi:10.1371/journal.pone.0027310.
934
Schloss, P.D., and Westcott, S.L., (2011). Assessing and improving methods
935
used in operational taxonomic unit-based approaches for 16S rRNA gene
936
sequence analysis. Appl. Environ. Microbiol. 77, 3219 – 3226.
937
Schrempf, H. (2006) “The Family Streptomycetaceae, Part II: Molecular Biology.” In
938
The prokaryotes : a handbook on the biology of bacteria Third Edition. eds.
939
Dworkin, M.,Falkow, S., Rosenberg, E., Schleifer, K.-H., Stackebrandt, E.M. (New
940
York NY : Springer), 3,605–622.
941
In review

37
Shannon, C. E. (1948). A mathematical theory of communication. Bell Sys.
942
Tech. J., 27, 379-423 and 623-656.
943
Simpson, E. H. (1949). Measurement of diversity. Nature 163, 688.
944
Soininen, J. (2012). Macroecology of unicellular organisms – patterns and
945
processes. Environ. Microbiol. Rep. 4(1), 10–22.
946
Souza-Egipsy, V., Aguilera, A., Mateo-Martí, E., Martín-Gago, J.A., and
947
Amils R. (2010). Fossilization of acidophilic microorganisms.
948
Geomicrobiol. J. 27, 692-706.
949
Spilde, M.N., Northup, D., Boston, P.J., Schelble, R.T., Dano, K.E., Crossey,
950
L.J., et al. (2005). Geomicrobiology of cave ferromanganese deposits: A
951
field and laboratory investigation. Geomicrobiol. J. 22, 99-116.
952
Stackebrandt, E., Rainey, F.A., and Ward-Rainey, N.L. (1997). Proposal for a
953
new hierarchic classification system, Actinobacteria classis nov. Int. J. Syst.
954
Micr. 47 (2): 479–491. doi:10.1099/00207713-47-2-479.
955
Steinert, G., Taylor, M.W., and Schupp, P.J. (2015). Diversity of
956
Actinobacteria associated with the marine ascidian Eudistoma toealensi.
957
Mar. Biotechnol. DOI: 10.1007/s10126-015-9622-3.
958
Subramani, R. and Aalbersberg, W. (2013). Culturable rare Actinomycetes:
959
diversity, isolation and marine natural product discovery. Appl. Microbiol.
960
Biot. 97, 9291-9321.
961
Tiwari, K. and Gupta, R.K. (2012) Rare actinomycetes: a potential storehouse
962
for novel antibiotics. Crit. Rev. Biotechnol. 32(2), 108–132.
963
Tiwari, K. and Gupta, R.K. (2013). Diversity and isolation of rare
964
actinomycetes: an overview. Crit. Rev. Microbiol. 39, 257-294.
965
Valverde, A., Tuffin, M., and Cowan, D.A. (2012). Biogeography of bacterial
966
communities in hot springs: a focus on the actinobacteria. Extremophiles
967
16, 669–679.
968
Velikonja, B.H., Tkavc, R., and Pašić, L. (2014). Diversity of cultivable
969
bacteria involved in the formation of macroscopic microbial colonies (cave
970
silver) on the walls of a cave in Slovenia. Int. J. Speleol. 43, 45-56.
971
Wawrik, B., Kutliev, D., Abdivasievna U.A., Kukor, J.J., Zylstra, G.J., and
972
Kerkhof, L. (2007). Biogeography of Actinomycete Communities and Type
973
II Polyketide Synthase Genes in Soils Collected in New Jersey and Central
974
Asia. Appl. Environ. Microbiol.73, 2982–2989.
975
Wu, Y., Tan, L., Liu, W., Wang, B., Wang, J., Cai, Y., et al. (2015). Profiling
976
bacterial diversity in a limestone cave of the western Loess Plateau of
977
China. Front. Microbiol. 6, 244.
978
Zhou, J.P., Gu, .Q., Zou, C.S., and Mo, M.H. (2007). Phylogenetic Diversity
979
of Bacteria in an Earth-Cave in Guizhou Province, Southwest of China. J.
980
Microbiol. 45(2),105-112.
981
982
In review

38
Figures and Tables:
983
984
985
986
Figure 1: A) World map of volcanic caves studied in this work: 1 – Hawaiian
987
volcanic caves (U.S.A.); 2 – Helmcken Falls Cave, British Columbia (Canada); 3 – New
988
In review

39
Mexico volcanic caves (U.S.A.); 4 – Azorean lava caves (Portugal); 5 – La Palma caves,
989
Canary Islands (Spain). B) General view of extensive yellow microbial mats from Gruta
990
da Terra Mole (Azores, Portugal). C) General view of white colonies forming dendritic
991
branches on basaltic lava from Fuente de la Canaria cave (La Palma Island, Spain). D)
992
Close-up view of a yellow colony from Gruta dos Montanheiros (Azores, Portugal). E)
993
Close-up view of white microbial mat covered with water droplets from Gruta da Terra
994
Mole (Azores, Portugal).
995
996
997
In review

40
998
In review

41
Figure 2: SEM images of colored microbial mats found in Azorean, Canadian,
999
Canarian and Hawaiian volcanic caves showing a large variety of microbial
1000
morphologies and spore surface ornamentation. A-B) Dense network of interwoven
1001
filaments in Honda del Bejenado and Fuente de la Canaria caves (La Palma Island,
1002
Spain); C) Dense masses of Streptomyces-like spore chains with hairy ornamentation
1003
from Cave 12 in El Malpais National Monument (New Mexico, U.S.A.); D) Coccoid
1004
cells with surface appendages or obtuse protuberances from Gruta da Terra Mole
1005
(Terceira Island, Azores, Portugal); E) Detailed view from (B) showing coccoid cells
1006
and clumps of spore chains with obtuse protuberances and surface appendages; F)
1007
Close-up view of clusters of Streptomyces-like spore chains with extensive hairy
1008
ornamentation from Gruta da Terra Mole (Terceira Island, Azores); G) Aggregates of
1009
coccoid cells with smooth surface and spherical cells arranged in chains resembling
1010
beads-on-a-string (arrow) from Bird Park Cave (Hawai'i, U.S.A.); H) Chain of
1011
Streptomyces-like arthrospores from Honda del Bejenado Cave (La Palma Island,
1012
Spain); I) Spores with spiny ornamentation from Helmcken Falls Cave, (British
1013
Columbia, Canada); J) Spiral spore chains of Streptomyces and a coccoid cell with
1014
obtuse protuberances (arrow) from Falda de La Horqueta Cave (La Palma Island,
1015
Spain); K) Large spheres with lumpy surface or protuberances connected by a network
1016
of filaments or appendages from Gruta dos Montanheiros (Pico Island, Azores); L)
1017
CaCO3 spheres coated with a filamentous network from the Tapa Section of the Kipuka
1018
Kanohina Cave Preserve (Hawai'i, U.S.A.).
1019
1020
In review

42
1021
Figure 3: SEM images of biosignatures found associated with microbial mats in
1022
Azorean, Canadian, Canarian and Hawaiian volcanic caves. A) Cell-shaped etched pits
1023
on mineral grain (arrow) from Helmcken Falls Cave (British Columbia Canada); B)
1024
Cell imprints (white arrow) and rods on EPS matrix from a white microbial mat in
1025
Gruta da Terra Mole (Terceira Island, Azores, Portugal); C) Microborings produced by
1026
euendolithic cells on silicified mineral grains from Ana Heva cave in Easter Island,
1027
Chile (adapted from Miller et al. 2014); D) Tubular mineralized sheaths embedded in
1028
EPS found on black deposits from Cueva del Llano de los Caños cave(La Palma Island,
1029
Spain). E) Reticulated filaments found in white microbial mats in the Kula Kai Caverns
1030
of the Kipuka Kanohina Cave Preserve (Hawai'i, U.S.A.); F) Close-up view of mineral
1031
encrusted filaments with reticulated ornamentation associated with white microbial
1032
mats on ooze-like deposits from Fuente de la Canaria cave (La Palma Island, Spain).
1033
In review

43
1034
1035
Figure 4. Order-Level delineation of the 16S rRNA gene sequences (A) and OTUs (B)
1036
in Azores, New Mexico and Hawai'i lava caves.
1037
In review

44
1038
Figure 5. Unrooted approximate maximum likelihood tree showing the
1039
relationship and occurrences of Actinobacteria OTUs across all four sample locations.
1040
Bootstrap values are indicated. A larger version of this figure is available in the
1041
supplemental material.
1042
1043
1044
In review

45
1045
Figure 6. Most abundant OTUs for Hawai’I, New Mexico and Azores. OTUs were
1046
clustered at 97% similarity and represent more than 1% of the total sequences
1047
recovered in each geographical location.
1048
In review

46
1049
Figure 7. Sequence identity based on BLAST comparisons to GenBank.
1050
1051
In review

47
1052
Figure 8. Order-Level delineation of the 16S rRNA gene sequences (A) and OTUs (B)
1053
in Azorean volcanic caves obtained by amplicon library.
1054
1055
1056
Table 1a. Summary of the observed richness, diversity indices, coverage and
1057
Chao 1 richness estimator at 97% similarity level at the three locations under study.
1058
Azores
New Mexico
Hawai'i
Richness
29.19
38.52
30
shannon
2.01 (1.73-2.29)
2.16 (1.86-2.46)
2.13 (1.87-2.39)
npshannon
2.35
2.69
2.48
simpson
0.31 (0.22-0.39)
0.28 (0.20-0.35)
0.22 (0.17-0.27)
invsimpson
3.31 (2.60-4.56)
3.62 (2.86-4.93)
4.49 (3.68-5.78)
In review

48
chao
97.79 (51.97-240.21)
158.98 (82.10-374.89)
100 (51.51-257.75)
coverage
0.86
0.78
0.85
1059
Table 1b. Summary of the characteristics of the pyrosequencing data. i.e.
1060
observed richness, diversity indices, coverage and Chao 1 richness estimator at 97%
1061
similarity level at the two islands from the Azorean archipelago under study.
1062
Azores
Pico
Terceira
Richness
382
191
141.57
shannon
1.99(1.96-2.02)
2. 67 (2.60-2.75)
1.68 (1.61-1.75)
npshannon
2.04
2.76
1.79
simpson
0.40(0.40-0.41)
0.24 (0.23-0.26)
0.45 (0.43-0.46)
invsimpson
2.48(2.43-2.53)
4.11 (3.88-4.37)
2.25 (2.15-2.35)
chao
529.70(477.26-611.02)
262.19 (229.43-322.89)
256.86 (203.65-355.78)
coverage
0.99
0.98
0.98
1063
1064
In review

49
Supplemental Figure 1. Venn diagrams representing shared OTU diversity of
1065
bacteria locations (97% sequence similarity). (A) Observed OUT; (B) Chao1 richness
1066
estimator.
1067
1068
Supplemental Table 1. Location and number of samples and sequences from
1069
each cave.
1070
Clone libraries
Amplicon libraries
Location
Cave
Number of
samples
Number of
sequences
Number of
samples
Number of
sequences
Hawai’i
Bird Park
3
61
-
-
Epperson’s
1
19
-
-
Kaumana
3
30
-
-
Maelstrom
1
2
-
-
Thurston
1
2
-
-
Kula Kai caverns
2
25
-
-
New Mexico
Cave 12
2
19
-
-
Cave 255
1
56
-
-
Cave 266
3
17
-
-
Cave 261
3
44
-
-
Cave 315
3
77
-
-
Pico
Furna do Lemos
3
62
2
395
Gruta dos
Montanheiros
1
4
2
521
Gruta da Ribeira do
Fundo
4
63
2
1043
Gruta das Torre
4
110
4
1562
Terceira
Algar do Carvão
3
41
2
-
In review

50
Gruta das Agulhas
3
25
2
1137
Gruta da Achada
3
52
2
1600
Gruta dos Buracos
3
39
2
3477
Gruta dos Balcões
3
43
2
1601
Gruta da Branca
Opala
4
107
2
1846
Gruta da Madre de
Deus
3
101
2
3800
Gruta do Natal
3
23
2
788
Gruta da Terra Mole
4
32
2
985
Gruta dos
Principiantes
4
66
2
103
Gruta da Malha
5
56
2
618
1071
Supplemental Table 1: GenBank Accession Numbers.
1072
Location
Accession Numbers
Hawai’i a
HM063012, HM063014-HM063020, HM063025, HM063027,
HM444834, HM444836, HM444862, HM444887, HM444891,
HM444892, HM444907, HM444912, HM444917, HM444919,
HM444931, HM444932, HM444938, HM444947, M444948,
HM444956, HM445516-HM445518, HM445523, HM445540,
HM445549, HM445551, HM445559-HM445561, HM445565,
KC569805, KC569806, KC569822, KC569827, KC569846,
KC569875, KC569892, KC569894, KT167192, KT167193
New Mexico
KC331621-KC331653, KC331655-KC331692, KC331694-
KC331743, KC331745-KC331751, KC331753-KC331838
In review

51
Azores
(Terceira) a
HM444993, HM444996, HM444998, HM445002, HM445004,
HM445009, HM445017-HM445019, HM445029, HM445047,
HM445082, HM445085, HM445102, HM445106, HM445113,
HM445132, HM445148, HM445185, HM445192, HM445223,
HM445238, HM445249, HM445251, HM445254, HM445255,
HM445265, HM445268, HM445285, HM445326, HM445368,
HM445393, HM445436, HM445437, HM445445, HM445486,
JF265974, JF265979, JF265985, JF265988, JF265999, JF266010,
JF266021, JF266027, JF266036, JF266037, JF266042, JF266065,
JF266068, JF266077, JF266079, JF266084, JF266085, JF266093,
JF266100, JF796752, JF796758, JF796763, JF796770, JF796771,
JF796776, JF796777, JF796781, JF796798, JN592625, JN592629,
JN592643, JN592647, JN592648, JN592652, JN592655,
JN592660, JN592670, JN592678, JN592689, JN592701,
JN592702, JN592709, JN600569, JN600572, JN600586,
JN600593, JN600597, JN606986,, JN606990, JN606992,
JN606998, JN607005, JN607006, JN607010, JN607015,
JN607016, JN607017, JN607024, JN607028, JN607030,
JN607033, JN607037, JN607041, JN607044, JN607045,
JN607051, JN607052, JN607066, JN607069, JN607080,
JN615637, JN615642, JN615645, JN615647, JN615657,
JN615666, JN615668, JN615673, JN615686, JN615690,
JN615700, JN615712, JN615726, JN615737, JN615745,
JN615750, JN615751, JN615780, JN615793, JN615802,
JN615805, JN615814, JN615818, JN615825, JN615828,
JN615830, JN615831, JN615837, JN615841, JN615844,
JN615851, JN615854, JN615855, JN615858, JN615859,
JN615861, JN615863, JN615865, JN615867, JN615870,
JN615873, JN615878, JN615880, JN615881, JN615888,
JN615889, JN615891, JN615892, JN615896, JN615897,
JN615899, JN615900, JN615901, JN615902, JN615906,
JN615907, JN615910, JN615913, JN615919, JN615923,
JN615925, JN615928, JN615933, JN615946, JN615951,
JN615954, JN615959, JN615960, JN615962, JN615963,
JN615965, JN615966, JN615971, JN615973, JN615977,
JN615979, JN615980, JN615981, JN615982, JN615985,
JN615988, JN615989, JN615994, JN615995, JN615996,
JN616002, JN616003, JN616005, JN616009, JN616011,
JN616012, JN616016, JN616018, JN616019, JN616020-
JN616022, JN616028, JN616030-JN616032, JN616034,
In review

52
JN616035, JN616040, JN616041, JN616044, JN616046,
JN616047, JN616056-JN616060, JN616063, JN616064,
JN616068, JN616073, JN616077, JN616079-JN616083,
JN616085, JN616087, JN616089, JN616091, JN616095,
JN616096, JN616102, JN616106-JN616108, JN616110,
JN616115, JN616116, JN616120, JN616124, JN616133-
JN616135, JN616140, JN616142, JN616143, JN616146,
JN616147, JN616149, JN616156, JN616157, JN616160,
JN616172, JN616178, JN616179, JN616181-JN616183,
JN616185, JN616188, JN616191, JN616194, JN616205,
JN616209, JN616211, JN616215, JN616219, JN616222,
JN616223, JN616224, JN616227, JN616231-JN616233,
JN616237-JN616239, JN643002, JN643009, JN643019,
JN672040, JN672045-JN672047, JN672049, JN672050,
JN672053, JN672061, JN672065, JN672066, JN672068,
JN672075-JN672077, JN672080, JN672082, JN672088,
JN672090, JN672091, JN672095, JN672096, JN672098,
JN672103, JN672105, JN672108, JN672109, JN672110-
JN672113, JN672116, JN672118-JN672120, JN672122,
JN672125, JN672126, JN672142, JN672148, JN672160,
JN672184, JN672197, JN672202, JN672205, JN672207,
JN672208, JN672216-JN672218, JN672220, JN672222,
JN672223, JN672224, JN672225, JN672227-JN672230,
JN672233, JN672236, JN672237, JN672239, JN672241,
JN672242, JN672251, JN672254, JN672257, JN672258,
JN672263, JN672265, JN672268, JN672280, JN672283,
JN672292, JN672293, JN672296, JN672311, JN672320,
JN672332, JN672339, JN672344, JN672358, JN672378,
JN672400, JN672416, JN672420, JN672486, JN672520,
JN672532, JN672543, JN672547, JN701036, JN701037,
JN701041, JN701043-JN701046, JN701051, JN701056,
JN701058, JN701060, JN701061, JN701064, JN701065,
JN701067, JN701072, JN701075, JN701080, JN701084,
JN701093, JN701094, JN701096, JN701099, JN701104-
JN701109, JN701112, JN701114, JN701116-JN701120,
JN701124, JN701125, JN701127, JN701131, JN701133,
JN701135-JN701139, JN701141, JN701143, JN701144,
JN701146-JN701149, JN701154, JN701156, JN701157,
JN701159-JN701162, JN701164, JN701166, JN701168,
JN701171, JN701173, JN701174, JN701176, JN701178,
In review

53
JN850065, JN850066, JN850067, JN850069, JN850070,
JN850071, JN850073, JN850074, JN850076, JN850078-
JN850081, JN850084, JN850086, JN850090, JN850091,
JN850093, JN850095, JN850096, JN850099-JN850111,
JN850114-JN850118, JN850120, JN850129, JN850150,
JN850173, JN850176-JN850179, JN850182, JN850186,
JN850188, JN850189, JN850191, JN850192, JN850194,
JN850195, JN850198, JN850199, JN850202-JN850204,
JN850207, JN850208, JN850210, JN850211, JN850213,
JN850218, JN850223, JN850227, JN850233, JN850237,
JN850241, JN850243, JN850244, JN850246, JN850250,
JN850259, JN850261, JN850270, JN850271, JN850273,
JN850280, JN850281, JN850284, JN850287, JN850288,
JN850291, JN850293, JN850294, JN850295, JN850299,
JN850302, JN850306, JN850307, JN850310, JN850311,
JN850313-JN850315, JN850317, JN850319-JN850323,
JN850325, JN850327-JN850330, JN850332-JN850338,
JN850340-JN850342, JN850344-JN850352, JN850354-
JN850358, JN850360-JN850362, JN850376, JN850419,
JN850440, JN850442, JN850483, JN850517, JN850520
Azores (Pico)
HM749656, HM749661, HM749663, HM749669, HM749671,
HM749672, HM749674, HM749681, HM749683, HM749688,
HM749689, HM749691, HM749697, HM749702, HM749703,
HM749705, HM749707, HM749709, HM749710, HM749714,
HM749723, HM749726, HM749734, HQ721104, HQ721105,
HQ721111, HQ721114, HQ721115, HQ721120, HQ721127,
HQ721137, HQ721147, HQ721148, HQ721152, HQ721154,
HQ721155, HQ721161-HQ721163, HQ721168, HQ721169,
HQ721170, HQ721172, JF265703, JF265714-JF265716,
JF265718, JF265739, JF265742, JF265744, JF265825, JF265833,
JF265843, JF265855, JF266105, JF266109, JF266113, JF266116,
JF266120, JF266123, JF266124, JF266126, JF266127, JF266133,
JF266146, JF266147, JF266161, JF266166, JF266168, JF266169,
JF266171, JF266172, JF266174, JF266175, JF266179, JF266184,
JF266187, JF266190, JF266196-JF266198, JF266200, JF266204 ,
JF266210, JF266229, JF266233, JF266236, JF266237, JF266238,
JF266240, JF266249, JF266254, JF266255, JF266275, JF266280,
JF266284, JF266287, JF266290, JF266292, JF266294, JF266298,
JF266299, JF266301, JF266303, JF266309, JF266316, JF266323,
In review

54
JF266326, JF266329, JF266341, JF266343, JF266344, JF266346-
JF266348, JF266350, JF266355, JF266360, JF266361, JF266365,
JF266366, JF266369, JF266373-JF266381, JF266386, JF266388-
JF266390, JF266393, JF266395, JF266396, JF266397, JF266400,
JF266401-JF266403, JF266407, JF266410, JF266413, JF266416,
JF266418, JF266419, JF266432, JF266448, JF266471, JF266483,
JF266495, JF266523, JF266534, JF266540, JF266555, JN696293,
JN696295, JN696301, JN696304, JN696305, JN696308,
JN696311, JN698909, JN698912, JN698913, JN698915,
JN698916, JN698917, JN698918, JN698919, JN698914,
JN698920, JN698921-JN698925, JN698927-JN698935,
JN698937-JN698942, JN698945, JN698947, JN801072,
JN801073, JN801078, JN801084-JN801086, JN801088-
JN801090,JN801095-JN801101, JN801103, JN801105,
JN801107, JN801111, JN801113-JN801117, JN801122,
JN801124,JN801126-JN801128, JN801130, JN801131,
JN801133, JN801135, JN801136, JN802348-JN802350,
JN802357, JN802363, JN802365, JN802367
Azores 454-
amplicons
KT230870-KT231775
a: The following accession numbers represent more than one sequence. HM444862 (2
1073
sequences), HM444891 (41), HM444892 (2), HM444919 (8), HM444948 (13),
1074
HM445517 (25), HM445518 (2), HM445523 (2), HM445559 (3), HM445561 (2),
1075
JN850250 (2).
1076
1077
In review

Figure 1.TIF
In review

Figure 2.JPEG
In review

Figure 3.JPEG
In review

Figure 4.JPEG
In review

Figure 5.JPEG
In review

Figure 6.JPEG
In review

Figure 7.JPEG
In review

Figure 8.TIF
In review