University of Southern California Dissertations and Theses

Microbial communities in marine sediments affecting and effecting biogeochemical cycling: influence of microbial ecology on geochemical transformations in two contrasting marine settings

(USC Thesis Other)

Microbial communities in marine sediments affecting and effecting biogeochemical cycling: influence of microbial ecology on geochemical transformations in two contrasting marine settings

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content 1
Microbial communities in marine sediments affecting and effecting
biogeochemical cycling:
Influence of microbial ecology on geochemical transformations in two contrasting
marine settings

Laura A. Zinke

A Dissertation Submitted to the Faculty of the
USC Graduate School
University of Southern California
In Partial Fulfillment of the Requirements for the
Degree Doctor of Philosophy
(BIOLOGICAL SCIENCES)

Department of Biological Sciences
Marine Biology and Biological Oceanography
August, 2018

Advisory Committee:
Jan P. Amend
Brandi Kiel Reese
John F. Heidelberg
Frank Corsetti

2
Acknowledgements

I would like to thank my advisors, Jan and Brandi, for seeing me through this dissertation
beginning to end, for giving me the freedom to make mistakes, and for forgiving my tendency to
fall asleep during the most inopportune times. I would like to thank Jan for giving a wide net to
explore, and Brandi for keeping me from wandering too far. This was a careful balance on your
parts, and I am immensely grateful.

I thank my committees, who taught me to be serious, to think both broadly and in detail, and
sometimes saw the glaringly obvious things I was too entrenched to notice. I would especially
like to thank Frank Corsetti for teaching me what a stromatolite really is, allowing me to attack
the Jug, and of course to keep my head buried in the sand. Literally. I also give special thanks to
John Heidelberg, who from the start of my PhD demonstrated that a scientist can be both
skeptical and excited, intensely knowledgeable, and always approachable.

I would like to thank my lab mates, including the wonderful undergraduates, including Juan
Orantes, Kimi Morales, Amanda Semler, Karla Abuyen, Alex Vasconcellos, Cynthia Bor, and
Emilie Skoog. Some of you I chose, some I inherited, some of you listened to my lab woes and
let me dominate the ‘clean’ hood. Our wonderful lab manager, Pratixa Savalia, thank you so
much for ordering my seemingless endless supply of taq and phenol, and for buying birthday
cakes and lab lunches.

Thank you to my office mates, who are my scientific sisters. Without the nights on the cheese
grater roof with Lily Momper, the years of swapping proofreading, figure advice, and sugary
treats with Guang-sin Lu, the brilliant photoshop skills of Jayme Fehylbus (I know what I said),
the bonding over silly dogs with Didi Bojanova, and the homebrewed magic from Heidi
Aronson, this dissertation would not have been possible.

I would like to thank my co-authors, both in the Amend lab and abroad, from whom I have
learned so much about science, collaboration, and fieldwork. Your generosity with samples, data,
writing, and insight have been instrumental. I also would like to thank those who have selflessly
helped with analyses and answered dozens of calls and emails that usually turned out to be
simple file path problems.

Finally, I would like to thank my family and friends. You have kept me mostly sane, and have
made graduate school more enjoyable than I imagined. Thank you, Robert Zinke, for always
responding to ‘coffee?’ I want to thank my parents for keeping me grounded and encouraged.
Lastly, I want to thank my husband, Jason Williams, who accepted my sneaky early morning
work sessions, but still made me occasionally leave the matrix and always made me laugh.

3
Table of Contents

Chapter 1: Introduction 5
Chapter 2: Thriving or Surviving? Evaluating active microbial guilds in
Baltic Sea sediment
12

Chapter 3. Climate influences microbial organic matter degradation in
Baltic Sea sediments

36
Chapter 4. Sediment microbial communities influenced by cool
hydrothermal fluid migration

80
Chapter 5. Microbial metabolism in low temperature hydrothermal
sediments at the Dorado Outcrop

123
Chapter 6. Conclusions 158

References 160
4

“The sea is everything. It covers seven-tenths of the terrestrial globe. Its breath is pure and life-
giving. It is an immense desert place where man is never lonely, for he senses the weaving of
Creation on every hand. It is the physical embodiment of a supernatural existence... For the sea is
itself nothing but love and emotion. It is the Living Infinite, as one of your poets has said. Nature
manifests herself in it, with her three kingdoms: mineral, vegetable, and animal. The ocean is the
vast reservoir of Nature.”
-Jules Verne, 20,000 Leagues Under the Sea

5
1. Introduction
Abstract
Sediments lay at the bottom of the ocean, containing the largest active reservoir of
organic carbon on this planet (Arndt et al., 2013). Microorganisms survive in these dark
environments, recycling and reworking organic matter over geologic timescales (Kallmeyer et
al., 2012; Orsi et al., 2018). These organisms are small, but an estimated 10
29
of them reside in
diverse subseafloor environments and collectively account for up to 22 Pg of carbon (Parkes et
al., 2014). As a whole, these communities drive subseafloor geochemical cycling, cause
diagenetic reactions, and alter carbon sequestration on Earth. Here, we investigate sediment
microbial communities in two marine settings: the Baltic Sea Basin and the Dorado Outcrop. The
location of the Baltic Sea rendered it susceptible to drastic environment changes during the
glacial cycles of the Pleistocene, resulting in the deposition of glacial tills and organic-poor
clays. By the mid-Holocene, however, the Baltic was an extremely productive marine-brackish
basin, and sediments from this time period are organic-rich. In 2013, the Integrated Ocean
Drilling Program (IODP) Expedition 347 recovered sediment up to 81 meters below the Baltic
seafloor. These cores represented an excellent opportunity to investigate 1) What are the active
microbial communities in the organic rich sediments; 2) How could these communities drive
organic matter remineralization; and 3) Are these communities related to depositional
conditions? Using metagenomics and metatranscriptomics, I found that the active microbial
communities were dominated by Atribacteria and Chloroflexi. Active metabolic processes
included methane consumption or production, fermentation, and reductive dehalogenation. There
was both genomic (metagenomics) and transcriptomic (metatranscriptomics) potential for
complex carbohydrate degradation, extracellular protein degradation, and various fermentation
6
pathways. In contrast to the Baltic, the Dorado Outcrop is a deep (3,000m water depth) basaltic
edifice in the Eastern Tropical Pacific Ocean. Dorado Outcrop is a site of Cool Hydrothermal
System (CHS) fluid flow, and in local sediments, CHS crustal fluid brings favorable terminal
electron acceptors including oxygen and nitrate into the sediment column from below. We
sampled sediment from across and near to the outcrop to determine: 1) What are the resident
microbial communities; 2) Do they differ between the background and the outcrop; and 3) What
are some of their metabolic functions? Sequencing of the 16S rRNA gene and metagenomics
revealed that though microbial communities are in many ways similar between nearby
‘background’ and Dorado Outcrop communities, they are distinct in the types and relative
percentages of many microbes, including Thaumarchaea and Chloroflexi. Nitrogen catabolism
was a dominant metabolic strategy amongst outcrop microbes, and ammonium and nitrite
oxidation were related to the outcrop conditions. In both the Baltic and Dorado, the mechanisms
by which microorganisms obtain energy and carbon in sediments highlighted the importance of
microbial communities in cycling of key elements, such as carbon and nitrogen, in the
subsurface.
Background
The marine subsurface
Subsurface life is ubiquitous. The first size estimate of the marine sediment biosphere
calculated that subseafloor hosted on the order of 10
30
microorganisms (Whitman et al., 1998).
Though more recent estimates have downsized this number (Kallmeyer et al., 2012; Parkes et al.,
2014), we nevertheless continue to find sediment-hosted life even far beneath the seafloor
(Inagaki et al., 2015), in organic-limited locations (D’Hondt, Spivack, Pockalny, Ferdelman,
7
Fischer, Kallmeyer, Abrams, Smith, Graham, Hasiuk, Schrum, and Stancin, 2009a), and under
extreme pressures (Bartlett, 2002).
Early studies of the marine subsurface often focused on continental shelf sediments (for
example, (Canfield et al., 1993; Thamdrup et al., 1994)), which are often formed through high
sedimentation rates and contain large amounts of organic matter (Hedges and Keil, 1995a).
Through these studies, it was discovered that microbial metabolism quickly drives sediment
anoxic, and eventually, to methane production conditions (as reviewed in (Jorgensen, 2006). In
response to these rather self-inflicted geochemical changes, microbial community succession
takes place through the sediment column (c.f.(Nealson, 1997). Traditionally, this succession is
from aerobic communities to nitrate and manganese, iron, and sulfate reducing communities,
until finally methanogenic and fermentative communities dominate. This zone of geochemical
and microbial succession has been well characterized in sediments worldwide (Sansone and
Martens, 1982; Canfield et al., 1993; Chen et al., 2017).
Since the 2000s, sediment samples from the open ocean have altered this viewpoint
(D'Hondt, 2004a; Roy et al., 2012; Ziebis, McManus, Ferdelman, Schmidt-Schierhorn, Bach,
Muratli, Edwards, and Villinger, 2012a). In oceanic deserts, sediments were found that contained
so little organic matter that they never went anaerobic (D 'Hondt et al., 2015), and oxygen usage
was shown deep into sediments (Roy et al., 2012). Sediments containing unexpected
geochemical transitions, such as multiple sulfate methane transition zones (D'Hondt, 2004a) or
increases in oxygen and nitrate (Ziebis, McManus, Ferdelman, Schmidt-Schierhorn, Bach,
Muratli, Edwards, and Villinger, 2012a), were discovered in many locations.
In the 2010s, commonly held ideas, such as that aerobic respiration and sulfate reduction
were the most important terminal electron acceptors in sediments, were questioned as new
8
measurements and models predicted the global dominance of other metabolisms, such as
methogenesis (Bowles et al., 2014). Advancement of geochemical and microbiological
techniques has allowed scientists to address increasing complex topics, such as cryptic elemental
cycling (Wasmund et al., 2017; Holmkvist et al.). New metabolisms have come to light (Dridi et
al., 2012), and long-known ones have been given new importance (Trembath-Reichert et al.,
2017).
The Baltic Sea Basin
The Baltic Sea Basin lies in Northern Europe, where it acts as a catchment basin for
almost 20% of Europe’s fresh water (Conley et al., 2009). In addition to these freshwater inputs,
saltwater enters the Baltic through the Danish Straits, bringing colder water from the North Sea
(T. Andrén, Björck, E. Andrén, Conley, Zillén, and Anjar, 2011a). This water mass juxtaposition
creates density stratification in the basin, leading to hypoxic or anoxic subbasins, and
contributing to fast (up to 500 cm per thousand years) sedimentation rates (Conley et al., 2009).
Recently deposited sediments in the Baltic tend to be methanogenic (e.g. (Beulig et al., 2018),
contain large percentages of organic matter (post cruise report), and harbor concentrated
microbial populations (post cruise report).
The Baltic has experienced markedly different climate stages in the recent past, however,
which are recorded in the basin sediments. The Scandinavian Ice Sheet (SIS) covered the basin
throughout the much of the Pleistocene (T. Andrén, Björck, E. Andrén, Conley, Zillén, and
Anjar, 2011a). Glacial retreat during the late Pleistocene and early Holocene interplayed with
isostatic rebound of Scandinavia, leading to turbulent conditions in the Baltic. At various times
in the past 20,000 years, the basin was a glacial lake, a freshwater lake, and a marine sea (T.
Andrén, Björck, E. Andrén, Conley, Zillén, and Anjar, 2011a). Variability in these conditions
9
can be related to variability in Baltic Sea Basin sediments. Sediments deposited during glacial
times tend to contain little organic matter and have low methane concentrations and salinities
(Egger et al.). In contrast, Baltic sediments associated with marine stages tend to contain
increased organic matter, large amounts of methane, and salinities closer to seawater values(T.
Andrén, Barker Jørgensen, et al., 2015).
Because of these shifts in depositional conditions and sediment types, the Baltic Sea
Basin has been labeled a ‘natural laboratory’ (Marshall, Karst, et al., 2017). In 2013, the
Integrated Ocean Drilling Program Expedition collected sediment samples from these various
sediment types. These sediments represented a chance to determine if the microbial assemblages
in the deep biosphere in the Baltic represent a unique deep biosphere community, or are the
result of initial sediment inoculum. Additionally, the variations in sources of organic matter
(marine vs terrestrial) and amounts (ranging between 0.3 and 8.65% found in IODP samples) in
these sediments provided natural settings for studying microbial remineralization or organics (T.
Andrén, Jorgensen, et al., 2015).

The Dorado Outcrop

Up to 25 million outcrops taller than 100 meters have been estimated to populate the
ocean floor (Wessel et al., 2010). These outcrops can facilitate fluid flow through the shallow
marine crusts, removing lithospheric heat and impact global elemental cycles (Fisher et al., 2003;
Harris et al., 2004; Wheat and Fisher, 2008). The Dorado Outcrop is part of a series of outcrops
which facilitate cool hydrothermal fluid flux through the shallow lithosphere. Dorado is
relatively small – only 150 m higher than surrounding seafloor (Wheat et al., 2017), which is far
below the 1.5 km seamounts detectable by satellite (Wessel et al., 2010). However, Dorado is
responsible for significant dis
10
charge (up to 20,000 L s
-1
) of oxic (54.5 µM), nitrate-rich (38 µM), and cool (~12.3°C) crustal
fluid(Wheat et al., 2017). In thin sediment patches on and immediately adjacent to Dorado
Outcrop, crustal fluid flux from the sediment-basement interface increases oxygen, nitrate, and
manganese oxide concentrations relative to background sediments.

Goals
Using the Baltic Sea Basin as a case study:
• Determine the composition and metabolism of the active microbial community in
sediments that are depleted in terminal electron acceptors.
• Understand how climate patterns affect long term carbon cycling potential in microbial
communities.
Using the Dorado Outcrop as a case study:
• Characterize microbial community response to sediments impacted by geochemical
conditions associated with CHS fluid flux.
• Describe the metabolic potential of CHS communities in sediments with delivery of
oxidants from underlying crust.
11
“With its untold depths, couldn't the sea keep alive such huge specimens of life from another
age, this sea that never changes while the land masses undergo almost continuous alteration?
Couldn't the heart of the ocean hide the last–remaining varieties of these titanic species, for
whom years are centuries and centuries millennia?”
– Jules Verne, 20,000 Leagues Under the Sea

12
2. Thriving or Surviving? Evaluating active microbial guilds in Baltic Sea
sediment
Adapted from the original publication in Environmental Microbiology Reports in 2017, in
collaboration with Megan Mullis, Jordan T. Bird, Ian P.G. Marshall, Bo Barker Jørgensen, Karen
G. Lloyd, Jan P. Amend, and Brandi Kiel Reese.
Abstract
Microbial life in the deep subsurface biosphere is taxonomically and metabolically
diverse, but it is vigorously debated whether the resident organisms are thriving (metabolizing,
maintaining cellular integrity, and expressing division genes) or just surviving. As part of
Integrated Ocean Drilling Program (IODP) Expedition 347: Baltic Sea Paleoenvironment, I
extracted and sequenced RNA from organic carbon-rich, nutrient-replete, and permanently
anoxic sediment. In stark contrast to the oligotrophic subsurface biosphere, Baltic Sea Basin
samples provided a unique opportunity to understand the balance between metabolism and other
cellular processes. Targeted sequencing of 16S rRNA transcripts showed Atribacteria (an
uncultured phylum) and Chloroflexi to be among the dominant and the active members of the
community. Metatranscriptomic analysis identified methane cycling, sulfur cycling, and
halogenated compound utilization as active in situ respiratory metabolisms. Genes for cellular
maintenance, cellular division, motility, and antimicrobial production were also transcribed. This
indicates that microbial life in deep subsurface Baltic Sea Basin sediments was not only alive,
but thriving.

13
Introduction
Marine sediments are the largest reservoir for organic matter (Hedges and Keil, 1995b;
Arndt et al., 2013) that may harbor as many microbial cells as the global ocean (Kallmeyer et al.,
2012; Parkes et al., 2014). Over the past several decades, numerous studies have described the
microbial communities of marine sediments using 16S gene sequencing, and more recently,
metagenomics and single-cell genomics ((Parkes et al., 1994; Reed et al., 2002; Biddle et al., 2006;
Inagaki, Nunoura, Nakagawa, Teske, Lever, Lauer, Suzuki, Takai, Delwiche, Colwell, Nealson,
Horikoshi, D’Hondt, et al., 2006; Lloyd et al., 2013; Orcutt, LaRowe, et al., 2013)). However, far fewer
analyses have demonstrated in situ activities of subsurface microbes, and many of these are from
basalt (Lever et al., 2013; Robador et al., 2014) or based on modeling (Roy et al., 2012). These
studies indicate that diverse metabolic strategies support microbes in various environments,
including aerobic respiration in sediments underlying gyres, sulfate reduction in subsurface
fluids, and methanogenesis in organic-rich sediments.
In 2014, sediments were recovered from the Baltic Sea Basin (BSB) via the Integrated
Ocean Drilling Program (IODP) and the European Consortium for Ocean Research Drilling
(ECORD) in order to understand the effect of depositional conditions (e.g., glacial lacustrine
versus non-glacial) on current microbial communities. Often, organic matter availability limits
growth in the oligotrophic subsurface biosphere (D'Hondt, 2004b; Jorgensen and Boetius, 2007;
Kobayashi et al., 2008; Dang et al., 2009; Hoehler and Jorgensen, 2013). However, compared to many
other subsurface sites, the BSB sediments are extremely rich in organic matter and are
potentially more representative of coastal organic-rich sites. In BSB surface sediments (less than
1 meter below seafloor), diverse microbial assemblages with activities related to methane
cycling, nitrate reduction, and sulfate reduction have been observed (Thureborn et al., 2013; 2016;
14
Reyes et al., 2017). However, the deeper, anoxic sediments, where many canonical electron
acceptors (e.g., O
2
, NO
x
, SO
4
2-
) are depleted, have not been similarly characterized.
Microbial activity in the deep biosphere has been inferred from cultivation-based
approaches, stable isotope fractionations, laboratory incubations, and metagenomic analyses.
Long doubling times, “bottle effects”, and resistance to isolation complicate cultivation-based
and incubation approaches (Orcutt, LaRowe, et al., 2013). Abiotic isotopic fractionation can
overprint biological fractionation (Shanks, 2001; Sim et al., 2011). Traditional metagenomic
analyses are often hindered by the presence of dead or dormant microbes, unexpressed genes,
and extracellular DNA preserved in organic-rich sediments (Corinaldesi et al., 2011; Torti et al.,
2015). However, DNA transcription has been shown to directly correlate with cellular activity
(Jansson et al., 2012). As such, RNA-based metatranscriptomic approaches have been employed
as an activity proxy in several natural and medical environments (Gilbert et al., 2008; Gosalbes
et al., 2011; Mason et al., 2012), and more recently also in marine sediments (Orsi et al., 2013;
(Urich et al., 2013; Pachiadaki et al., 2016). Here, I used metatranscriptomics and targeted
quantification of 16S rRNA to examine the active microbial community abundance, structure,
and function of deeply buried sediments from IODP Expedition 347: Baltic Sea
Paleoenvironment. Compared to other deep marine sediments studied, these Baltic Sea sediments
are young (Holocene in age), sulfate-depleted, and organic matter rich. These sediments
represent an opportunity to characterize active microbial community function in a nutrient
replete deep biosphere setting.

15
Methods
Sample collection
Samples were collected by the Integrated Ocean Drilling Program on board the MPSV
Greatship Manisha during September-November, 2013. Cores were collected via advanced
piston coring in 5-6 m sections and further subsectioned into 1.5 m sections. Prior to drilling
activities, perfluorocarbon tracer (PFT) was added to the drill fluid to track the amount of
contamination of the core due to drilling. Microbiological cores were assessed for visible signs
of disturbance, including cracks in the core, bubbles, or liquid within the core. Cores were
sectioned into whole round cores in a 12°C microbiological container using hacksaws for the
core liner and sterilized wire for the sediment. Sediment cores for nucleic acid analyses were
immediately frozen at -80°C on ship and shipped to the laboratory on dry ice (Andren et al.,
2015).
RNA extraction
Sediment cores for microbiological analyses were selected based on geochemical context,
lithology, and lack of potential contamination based on PFT data. In the cores here, PFT tracer
measurements indicated contaminating microbes from drill fluid would be a maximum of
0.001% of the total counted microbial community in sample 59E-42m and 0.0003% in 63E-12m.
Sample 59E-15m showed no PFC tracer in the core interior.
Frozen sediment was chipped from whole round cores in a dedicated HEPA-filtered clean
room at Texas A&M University - Corpus Christi. All instruments were flame-sterilized and
edges of the sediment core were avoided. RNA was extracted from sediment using the MoBio
PowerSoil RNA kit (MoBio Laboratories, Carlsbad, CA) following manufacturer’s instructions.
Approximately 7.5 g of sediment from each sample was extracted, which was divided over three
16
reactions (2.5 g each). All RNA extraction steps were performed in a clean hood sterilized with
UV light and RNaseZap (Ambion, Foster City, CA) while using certified RNase-free
consumables and glassware baked at 450
o
C. Researchers wore face-masks and hair caps to avoid
sample contamination. The final RNA-containing pellets were sequentially combined in 50µl of
PCR-grade RNase-free water. No-template contamination controls were run alongside the
samples and were assessed for sterility through PCR amplification and agarose gel visualization.
Total RNA extractions were treated with Ambion Turbo DNase (ThermoFisher
Scientific, Waltham, MA) according to manufacturer protocols. Resulting RNA purity and
quantity was checked using the Eppendorf Biospectrometer (Eppendorf, Hauppauge, NY) and by
reverse transcription of the 16S rRNA gene transcript followed by PCR amplification of the
complimentary DNA (cDNA). DNase-treated RNA extract was PCR amplified to determine if
DNase treatment was effective. The PCR reaction was visualized on a 1% agarose gel and no
reaction was visible, indicating thorough DNase treatment. Extracts were shipped on dry ice to
the Molecular Research DNA Laboratory, LLC in Shallowater, Texas for library preparation and
sequencing analysis.
Droplet Digital PCR
The 16S SSU rRNA gene transcripts of Bacteria and Archaea were quantified using
droplet digital PCR (ddPCR). The extractions were first treated with DNase enzyme to remove
any contaminating DNA (Ambion, Foster City, CA). An aliquot was reverse transcribed by
mixing 2.5µl of template and 2.5µl of 5µM reverse primer, incubating at 70°C for 5 minutes,
then adding 20µl of mastermix containing M-MLV reverse transcriptase reaction buffer (final
concentration 50mM Tris-HCl (pH 8.3), 75mM KCl, 3mM MgCl2, and 10mM DTT), 250µM
dNTPs, and 200U M-MLV reverse transcriptase) (Promega, Madison, WI). The reverse primer
17
B518R (5’-GCTATTACCGCGGCTGCTGG-3’) (Nogales et al 1999) was used for Bacteria-
specific reactions. The reverse primer A519R (5’-GGTDTTACCGCGGCKGCTG-3’) (Wang
and Qian, 2011) was used for Archaea-specific reactions. cDNA was synthesized by incubating
the reaction for 1 hour at 37°C.
Each sample was analyzed in triplicate on the ddPCR using 1µl of cDNA per 25µl
reaction with the EvaGreen ddPCR reaction mixture (Bio-Rad Laboratories, Hercules, CA), and
100nM of each forward and reverse primer. Reverse primers were the same as above. Forward
primers were B341F (5’-CCTACGGGRSGCAGCAG-3’) targeting Bacteria and A340F (5’-
CCCTAYGGGGYGCASCAG-3’) targeting Archaea (Y. Wang and Qian, 2009). Positive
controls of known concentration and template-free control reactions were included in every set
of reactions. Reaction droplets were generated from 20µl of PCR mixture and 70µl of QX200
droplet generation oil using a Bio-Rad QX200 Droplet Generator (BioRad, Hercules, CA). The
droplet mixtures were transferred to a 96-well plate were PCR amplified using the EvaGreen
recommended cycling conditions: 95°C for 5 minutes, 40 cycles of 95°C for 30s and 60°C for
60s, reaction stabilization at 4°C for 5 min then 90°C for 5 min, and a final hold at 4°C. Droplets
were read for template amplification using a Bio-Rad QX200 Droplet Analyzer (BioRad,
Hercules, CA). Starting template concentrations were calculated using QuantaSoft Analysis Pro
Software in ABS (absolute) mode with a manually set threshold. Further analyses including
calculation of transcript abundance in sediment, determination of averages and standard
deviation, and RNA/DNA ratios were performed in Microsoft Excel®.
Library Preparation and Sequencing
For metatranscriptomes, complementary DNA synthesis was performed using the
QuantiTech Whole Transcriptome Amplification kit following manufacturer’s instructions
18
(Qiagen, Hilden, Germany). Libraries were generated from cDNA using the Nextera DNA
sample preparation kit (Illumina, San Diego, CA) and were sequenced on the Illumina HiSeq
2500 platform (Illumina, San Diego, CA) for 500 cycles with 250 bp paired-end chemistry.
The 16S rRNA transcript sequencing was performed using 454 Life Sciences GS FLX
sequencing platform and Titanium chemistry (Roche, Branford, CT, USA). Bacteria-specific
primers 28F (5'-GAGTTTGATCNTGGCTCAG-3') and 519R (5'-
GTNTTACNGCGGCKGCTG-3') (Reese et al., 2013) for Bacteria and 349F (5ʹ-
GYGCASCAGKCGMGAAW-3ʹ) (Raskin et al., 1994) and Archaea-specific 806R (5ʹ-
GGACTACVSGGGTATCTAAT-3’) (Takai and Horikoshi, 2000) were used. Samples were
reverse transcribed and PCR amplified with barcoded primers for 30 cycles (Dowd et al., 2008).
Amplicons were mixed in equal concentrations and purified with Agencourt Ampure Bead
(Agencourt Bioscience Corporation, MA, USA). Amplicons were sequenced using GS FLX
Titanium Sequencing Kit XLR70 (Roche, Branford, CT, USA) following manufacturer
protocols.
16S rRNA Bioinformatics
A total of 97,594 bacterial sequences were produced from all three samples, and 78,412
sequences passed quality control with an average read length of 370 bp and an average of 26,137
sequences per sample (Supplementary Table 2).
A total of 97,107 archaeal sequences were produced from the 16S rRNA transcript sequencing,
and 77,038 sequences passed quality control with an average length of 370 bp. A total of 77,038
archaeal 16S rRNA transcript target sequences passed quality control.
The 16S rRNA transcript sequences were processed in QIIME (Caporaso et al., 2010) using a
minimum quality score of 25, a minimum length of 200 and maximum length of 1000, no
19
ambiguous bases, and no primer mismatches. OTUs were clustered de novo at 97% similarity
using the pick_de_novo_otus.py command and the uclust program (Edgar, 2010). Taxonomy
was assigned to OTUs against the Silva V123 database (Quast et al., 2012; Yilmaz et al.,
2013)using the uclust algorithm with a minimum consensus fraction of 0.51 and minimum
percent similarity of 0.9 (Edgar, 2010). Sequences have been deposited to NCBI under
BioProject PRJNA388481.
Metatranscriptome Bioinformatics
A total of 50,315,718 read pairs were sequenced and 49,171,259 had both reads pass quality
control. Between 67 and 75% of the read pairs had overlap between the read pairs and were
merged. The total number of merged and unmerged reads annotated against the KEGG database
was 2,290,437. Data is reported here as number of reads assigned to a process or protein as a
percentage of all reads passing quality control (Supplementary Table 2).
Metatranscriptome reads were initially filtered in the FastQ toolkit using a minimum
quality score of 25, a minimum read length of 32, and minimum complexity of 25. Sequencing
adapters were trimmed in the FastQ toolkit (BaseSpace Labs, Illumina, Inc., San Diego, CA). All
pairs with both sequences that passed quality control were aligned against the Non-Redundant
(nr) NCBI database (downloaded 2014) using BLASTX mode in DIAMOND (Buchfink et al
2015). Function was assigned to aligned sequences in the MEtaGenome ANalyzer (MEGAN) V6
(Huson et al., 2007; 2011; Huson, Beier, Flade, Górska, El-Hadidi, Mitra, Ruscheweyh, and Tappu,
2016a) against the 2011 release of Kyoto Encyclopedia of Genes and Genomes (KEGG)
(Kanehisa et al., 2016) and the May 2015 SEED database (Overbeek et al., 2005). All
metatranscriptomes were normalized in MEGAN to an equal number of reads as the sample with
the lowest abundance. Percentages of transcripts assigned to genes were calculated by dividing
20
the number of reads assigned by the total number of reads and converting to percentage.
Biochemical and genetic pathways were determined based on KEGG pathways (referenced
online in 2017) and a review of primary literature. Sequence data was submitted to NCBI as
BioProject PRJNA388431.
Results
Site Description
The BSB is relatively shallow (average water depth 54 m), with a halocline due to
density-driven stratification from freshwater runoff present at a depth of 60-80 m. This causes
sub-basins to be seasonally or permanently hypoxic or anoxic (< 0.2 mg O
2
L
-1
) (Carstensen et
al., 2014). The combination of eutrophic conditions in the surface water, shallow average water
depth, and regional dysoxia have resulted in high sedimentation rates (0.1 – 0.5 cm year
-1
) across
the BSB over the past ~11.7 kyr (Andren et al., 2015). In the present study, I focus on two BSB
locations, the Landsort Deep and Little Belt (hereafter referred to as Sites M59 and M63,
respectively) (Fig 1a).
Site M59 (55°0.29ʹN, 10°6.49ʹE) is located at an entryway where marine waters from the
North Sea flow into the Baltic Sea (Fig 1a). The top 46.2 m of sediment at this site was classified
as biosiliceous clay deposited during marine to brackish phases of the Baltic Sea Basin (Andren
et al., 2015), with TOC levels at 4.3-7.4 percent dry weight (wt%). Chloride-based salinity in the
porewater is consistently high (23.75-24.57) in the top 15 m, and then decreases to a minimum of
7.0 by 65.38 meters below seafloor (mbsf). The shape of the porewater salinity profile indicated
post-depositional diffusion of ions from the marine-influenced Holocene sediments to the
underlying lacustrine sediments (Supplementary Table 1).
At 459 m water depth, the Site 63E (58°37.34ʹN, 18°15.25ʹE) is in the deepest part of the
21
BSB. It experiences periodic anoxia that occasionally results in ferruginous or euxinic conditions
(Hardisty et al., 2016). The sediments at Site M63 were characterized by organic-rich clays.
Nanofossils and porewater salinity indicate that sediment deposition in the top 25 m occurred
during a marine-brackish phase (Andren et al., 2015). The highest salinity (12.5) was measured
at 14.95 mbsf and decreased to <1 in the glaciolacustrine sediments below 50 mbsf. The total
organic carbon (TOC) was 6.41 wt% in the marine sediment, and <1 wt% in the glaciolacustrine
sediments (Supplementary Table 1).
Bacteria to Archaea ratios
Using droplet digital PCR, I determined the relative abundance of archaeal and bacterial
16S rRNA transcript copy numbers. Archaea comprised 2.7-4.7% of the total rRNA quantified
and, correspondingly, Bacteria comprised 97.3-95.3%, with an average Archaea to Bacteria ratio
of 1:26 (Supplementary Table 2).
Microbial Diversity
RNA was extracted from two depths (15 and 42 mbsf) at Site M59 and one depth (12
mbsf) at Site M63, hereafter referred to as 59E-15m, 59E-42m, and 63E-12m, respectively. The
16S rRNA gene transcripts were sequenced using archaeal and bacterial specific primers, and
subsequently classified at the 80% confidence interval against the Silva 123 database (Pruesse et
al., 2007). The relative abundances of the gene transcripts that could be classified are reported
(Fig. 1b).
Atribacteria dominated at 63E-12m (76.6% of classifiable rRNA transcript sequences)
and at 59E-42m (59.1%), but they were less abundant at 59E-15m (21.7%) (Fig. 1b). The
Chloroflexi were also abundant at 59E-15m (26.9%) and at 59E-42m (23.9%), but less abundant
at 63E-12m (6.1%). Most of the Chloroflexi were further assigned to the genus Dehalococcoides,
22
which has cultivated isolates proven capable of reductive dehalogenation (Müller et al., 2004;
Loffler et al., 2013) but other Chloroflexi taxa were present as well. Other bacterial phyla
identified in the 16S rRNA transcripts included, for example, Aminicenantes (6.7% at 59E-15m),
Planctomycetes (5.9% at 59E-15m), and Proteobacteria.
Within the Archaea, the transcripts annotated as Euryarchaeota were 22% at 59E-15m,
26.3% at 63E-12m, and 87.5% at 59E-42m. Most of these sequences were further classified as
ANME-1b or as unclassified families within the order Methanomicrobiales. In sample 59E-15m,
18% of the archaeal sequences were related to Lokiarchaeota, previously referred to as the
Marine Benthic Group B (MBG-B) (Spang et al., 2015). The remaining archaeal rRNA
sequences were assigned to Thaumarchaeota, Bathyarchaeota, or could not be assigned at the
phylum level.
Active metabolisms
Metabolic pathways transcribed by the microbial communities were determined from
annotated sequences using the KEGG database (Supplementary Fig. 1, Supplementary Table 2
and Methods). An average of 89.9% of reads were not assigned to any protein in this database.
Of the remaining reads, more were assigned to the ‘metabolism’ category (3.72 - 4.9% of all
reads) than to ‘genetic information processing’ (0.87 - 1.32%), ‘environmental information
processing’ (0.67 - 1.04%), or ‘cellular processes’ (0.19 - 0.24%) (Supplementary Fig. 1). It
should be noted that the relative abundance of transcripts represents the number of transcripts
assigned to these processes relative to the whole of the metatranscriptome, and is not equivalent
to the total number of transcripts in situ. Within the ‘metabolism’ category, the active pathways
with the most gene expression included carbohydrate, energy, and amino acid metabolisms
(Supplementary Fig. 1). These categories are broad, but after subsampling the
23
metatranscriptomes to the smallest metatranscriptome in silico and normalizing to the abundance
of assigned transcripts, samples show notable differences. The deepest and presumably the oldest
sample, 59E-42m, contained the highest relative percentages of transcripts assigned to amino
acid metabolism (1.59% of all transcripts), energy metabolism (1.26%), and nucleotide
metabolism (0.91%).
Methanogenesis
Methane concentrations ranged from 2.92 mM near the surface to 2.02 mM at 47.88 m
mbsf at site M59A (Andren et al., 2015). At site M63, methane was detected at all depths down
to 72 m. Significant methane outgassed from the sediment cores during sampling, so the reported
concentrations represent minimum values (Egger et al.).
Transcripts for genes involved in methanogenesis were found in our metatranscriptomes.
These include genes that encode for methyl-CoM reductase (mcr), heterodisulfide reductases
(hdr), formylmethanofuran reductases (fwd), methylenetetrahydromethanopterin dehydrogenase
(mtd), and 5,10-methylenetetrahydromethanopterin reductase (mtr) (Fig. 2; Supplementary Table
3).
The absence of the fwd, ftr, mch, mtd, and hmd genes in the deepest sample (59E-42m)
indicates that hydrogenotrophic methanogenesis was not fully expressed there or that the
methanogenic community involved in this process was small and below detection. However,
transcripts for acetoclastic methanogenesis (acds) and methylotrophic methanogenesis (mta, mtb,
mtt), as well as subunits for mcr and other steps of methanogenesis were found in all three
sediment samples, including in 59E-42m.

24
Sulfate Reduction
Sulfate in sediment porewater may serve as a terminal electron acceptor in the oxidation
of organic matter by sulfate reducing bacteria. The porewater sulfate concentration at site M59
was at 1.44 mM at 0.73 m mbsf, but dropped to low micromolar levels at most depths down to
80 m mbsf, including in samples studied here. At site M63, sulfate was <0.01 mM at most depths
down to 74 m mbsf (Andren et al., 2015).
Most of the genes related to sulfur cycling were classified as assimilatory (e.g., cys,
paps), but a few dissimilatory transcripts were noted as well. Genes encoding proteins which
mediate the first steps of dissimilatory sulfate reduction to sulfite were expressed, including
sulfate adenylyltransferase (sat) and adenosine-5-phosphosulfate reductase (apr). These were
only found in sample 59E-15m (Fig. 3), which had the most potential seawater intrusion during
sampling. None of the transcripts from any of the samples were annotated as dissimilatory sulfite
reductase (dsr) against the KEGG database. However, when assigning function against the SEED
database, three transcripts from 59E-15m were annotated as dsrAB. In addition to mediating
sulfate reduction, the dsrAB protein has been posited to produce sulfate through sulfur oxidation
(Loy et al., 2009; Muller et al., 2015).
Polysulfide reductase genes (psr), which confer to microbes the ability to use
polysulfides as electron acceptors (Jormakka et al., 2008), were found in 59E-15m. In sediments
where sulfides are produced during diagenesis, expression of the psr gene could indicate some
microbes are polysulfide reducers. These data provide evidence that a subset of the active
microbial community could be using polysulfide reduction for energy in sulfate-depleted
environment.

25
Reductive dehalogenation
Consistent with this, a canonical reductive dehalogenation gene, trichloroethylene
reductive dehalogenase (tceA), was expressed in all three samples. However, the relative
proportion of transcripts for this gene was higher at site M59E, which featured higher seawater-
influenced

chloride

concentrations than site M63E. Expressed transcripts were observed for
genes in pathways for degradation of other chloroalkanes, including
carboxymethylenebutanolidase (CMBL) and nitrophenyl phosphatase (NPPase) in 63E-12m, and
haloacid dehalogenase (DEX) in both 63E-12m and 59E-15m (Fig 2), though complete pathways
were not found.
Active cellular processes
In the three samples examined in the current study, there was evidence of motility gene
transcription (KEGG category “Flagellar assembly”). In addition, transcripts of flagellar
biosynthesis genes (e.g., fla, flg, flh, and fli) were present in all samples (Fig 3; Supplementary
Table 4). Genes associated with gliding and twitching motility (pil, gld, fts) were also transcribed
in the sediments in all depths. Reads assigned to mot and che proteins, which effect and regulate
flagellar movement, were identified in all samples, with a higher percentage at site 59E than at
63E-12m.
Transcripts were assigned to functional genes within the KEGG categories ‘cell growth
and death’ and ‘nucleotide metabolism’ (Supplementary Fig. 1), and multiple genes involved in
cell division were found within the metatranscriptomes of all three samples (Fig 3). In addition,
transcripts for the protein responsible for initiating cell division and recruiting other proteins to
produce a new cell wall (fts) were identified, as were ATP-dependent lon and clp proteases,
which are indicative of replicative DNA helicases, DNA primases, and DNA polymerases (Fig
26
4). To survive in the deep biosphere, cells must also maintain structural and DNA integrity
(Lomstein et al., 2012). The percent of reads assigned to DNA repair in the deepest sample (59E-
42m) was double the percent of reads assigned to DNA repair in the shallower samples.
In all of our BSB samples, gene transcripts for the PKS pathway were expressed,
including for ansamycin biosynthesis (e.g., tktAB), tetracycline biosynthesis (e.g., accABCD, and
bccP), polyketide sugar unit biosynthesis (e.g., rfbACD, rffGH, rmlCD), and vancomycin
biosynthesis (e.g., rfbB and rffG) (Supplementary Table 4).
Discussion
Community structure
I found a clear dominance of Bacteria over Archaea in the active communities. A
separate study on nearby sediment cores using quantitative PCR of 16S rRNA genes also showed
a bacterial dominance, but, compared to our findings, noted higher percentages of Archaea (26-
47% of rRNA gene copies) at one site (M59E) (Buongiorno et al., 2017) and a recent study on
sediments from nearby Aarhus Bay found higher percentages of Archaea in deeper sediments as
well (Chen et al., 2017). These data suggest that Bacteria dominate the active community, or
contain more ribosomes per cell, despite numerical abundance of Archaea in DNA-based
analyses.
Atribacteria or Chloroflexi, two groups that are ubiquitous in anoxic marine sediments
(Parkes et al., 2014), were the dominant active bacterial phyla in all three samples (Fig 1b). The
Atribacteria and Aminicenantes are bacterial phyla that have no cultured species. Within the
Archaea, many of the groups found were related to methane cycling lineages. For example, I
found 16S rRNA sequences assigned within the ANME-1b group, often observed co-occurring
with sulfate reducing bacteria (Girguis et al., 2003), are capable of the anaerobic oxidation of
27
methane (AOM), but also are shown to perform methanogenesis (Lloyd, Lapham, and Teske,
2006a; Lloyd et al., 2011). The Methanomicrobiales, an order of methanogens with members that
use H2/CO or formate as substrates (Garcia et al., 2006), were observed in all our samples and
corresponded with relatively high methane concentrations. Lokiarchaeota sequences were also
found in Baltic Sea sediments here. Sequences attributed to this candidate phylum has been
found in marine sediments, and Lokiarchaeota specifically have been described as a key
microbial group in the sedimentary deep biosphere (Starnawski et al., 2017). Overall, the active
microbes identified via 16S rRNA sequencing agree well with those identified as deep biosphere
persisting microbes (Starnawski et al., 2017).
Active microbial metabolism
The differences in relative abundances of broad metabolic categories indicates that the
deeper community allocates more transcriptional activity on maintenance and energy
conservation than do the shallower communities, which have younger and presumably more
labile substrates to respire. This agrees with a recent study by Kempes et al. (2017) in which
smaller, more energy starved Bacteria focus cellular energy on protein and nucleotide
metabolisms, and Orsi et al. (2015) which found that protein cycling accounts for a large percent
of cellular activity at low metabolic rates.
Methane metabolism
Pathways for methane production were found to be active in sediments here, which is
concordant with the mM concentrations of methane in situ. The abundance of methylated
compound usage gene transcripts indicates that methylated compound breakdown and acetate
utilization were important methanogenesis pathways in the BSB sediments (Fig. 2).
Methyltrophic methanogenesis is recognized as an important metabolic pathway in estuarine
28
sediments and in sulfate-reducing zones of marine sediments (Valentine, 2011). Additionally,
this process has been reported in several deep subseafloor environments, including the Peru
margin (Orsi et al., 2013). The combination of expressed methanogenesis protein-encoding
genes with millimolar concentrations of methane points to biological methane production as the
main terminal process of the microbial food chain at these sites.
It has been proposed that these genes are used in AOM, but only when running the
methanogenesis pathway in reverse (F.-P. Wang et al., 2013). The expression of these genes and
the presence of ANME ribosomal sequences could indicate an active AOM community in these
sediments, but with an unknown electron acceptor. It has been suggested that labile iron oxides
may play this role in Baltic sediments, including at site 63E (Egger et al., 2017).
Sulfur cycling
In an environment deposited under marine water conditions, as the sediments here were,
such low concentrations are the result of microbial sulfate reduction, which is sometimes
accompanied by detectable sulfide (Andren et al., 2015). However, low or undetectable levels of
sulfide—as was the case in these sediments—may be due to precipitation with iron, biological
oxidation, or chemical oxidation (Reese et al., 2013). There was little evidence here for
biologically mediated sulfur cycling in 59E-41m and 63E-11m, but there was however evidence
for sulfur cycling in 59E-15m. Sulfate reduction or production, indicated dsr transcripts, and
polysulfide reduction, indicated by psr transcripts, were apparently active in 59E-15m.
Potentially, sulfur cycling occurs even deep into the methanogenic zone.
Reductive dehalogenation
Halogenated organic compounds are produced naturally in marine environments and
some are subsequently buried in sediment (Häggblom and Bossert, 2003). Under anoxic
29
conditions, they can serve as terminal electron acceptors in biological reductive dehalogenation.
Labile carbon compounds are released as byproducts, which can serve as carbon and energy
sources for non-dehalogenating microbes (Futagami et al., 2009; Jorgensen and Marshall, 2016).
Elevated porewater concentrations of bromide relative to chloride indicate reductive
dehalogenation of brominated organic compounds (Berg and Solomon). The Br/Cl molar ratio
within the BSB sediment porewater was greater than that in seawater and increased with depth at
both sites. At site M59, this ratio was 1.67×10
-3
at the surface and increased to 4.06×10
-3
at 42
mbsf, and at site M63, the ratio was 1.78 ×10
-3
at the surface increasing to 2.08 ×10
-3
at 12.3
mbsf (Andren et al., 2015). This points to reductive dehalogenation as a plausible source of
energy within the BSB sediment. Reductive dehalogenation has been widely observed in
subsurface sediment from the Peru Margin, eastern equatorial Pacific, Juan de Fuca Ridge flank,
and northwest Pacific near Japan (Futagami et al., 2009). At those locations and in the BSB, this
metabolic process may play a significant role in biogeochemical cycles of chlorine, iodine,
bromine, and halogenated carbon substrates and facilitate organic matter degradation.
Active cellular processes
In static deep marine sediment that lack noteworthy lateral or vertical fluid flow, the
ability of a microbe to move through pore spaces to find nutrients might confer a useful, yet
costly advantage (Parkes et al., 2000). Motility and chemotaxis were recently estimated to be
perhaps too costly in the deep biosphere (Hoehler and Jorgensen, 2013), but the high nutrient
concentrations in BSB sediments could potentially support such activities. Additionally, motility
would be beneficial in accessing solid substrates such as particulate organic matter and mineral-
bound phosphates. The presence of motility genes here demonstrates that at some portion of the
active community was motile. Other subsurface studies have also found these genes expressed,
30
suggesting their presence in deep marine sediments(Orsi et al., 2013). In brief, it appears that at
least some deep subsurface microbes retain the capacity to move toward food or other substrates.
To survive, subseafloor microorganisms likely engage in a wide variety of interactions
with both the environment and each other. With increasing depth, metabolisms slow and growth
rates decrease, suggesting that cell maintenance becomes an increasingly important factor in
survival.

To respond to changing environmental conditions, microbial populations can 1)
outcompete their neighbors by adapting their metabolisms to the available resources, 2) enter
into a mutualistic cooperation with their neighbors, and/or 3) inhibit or kill their neighbors
through the production of secondary metabolites (i.e., antimicrobials) that confer a competitive
advantage. The two most common mechanisms of antimicrobial production are through
polyketide synthase (PKS) and nonribosomal polyketide synthase (NRPS) pathways (Winter et
al., 2011), both of which form specific peptides with conserved and variable domains, which
account for the diversity of natural antimicrobials (Clardy and Walsh, 2004; Yin et al., 2007).
Antimicrobial production genes have been isolated from a diverse set of microbial lineages and
environments, but most research to date has focused on Actinomycetes and terrestrial soil
environments. Given the increasing prevalence of antimicrobial resistance, further exploration
into natural products is warranted, especially in poorly-studied environments such as the deep
subsurface.
In the sedimentary biosphere, cell numbers decrease exponentially with depth, and it is
unclear if the resident communities are multiplying or only turning over biomass without
division (Lomstein et al., 2012). Extremely long doubling times of over 1000 years have been
estimated for deep subsurface microbial communities (Hoehler and Jorgensen, 2013). However,
the cell counts in BSB sediments are higher than those predicted from global regression lines
31
(Kallmeyer et al., 2012; Parkes et al., 2014; Buongiorno et al., 2017). Due the high
sedimentation rate in the BSB, even deeply buried sediment can be relatively young (less than
<10,000 years old) and the sediment organic matter is likely more labile than at other deep
biosphere sites. This not only appears to support microbial life, but also active cellular division.
This agrees with a recent study that calculated genetic mutation rates in the deep biosphere and
concluded that DNA mutation repair is maintained in subsurface sediments (Starnawski et al.,
2017), and further demonstrates the importance of maintaining current cellular integrity in the
deep biosphere. The active gene expression and maintenance of cellular integrity indicates that
the cells in BSB sediment were active and dividing, and what I define as thriving, and not just
dying slowly.
Summary
The microbial communities within the organic-rich, deeply-buried sediment of the Baltic
Sea are diverse and active. RNA-based analyses revealed that these communities carry out a
wide range of metabolisms, including methanogenesis, sulfate reduction, and reductive
dehalogenation. I found gene expression associated with cellular maintenance, division, and
motility, which could imply that microbes in the active community not only manage to survive in
these sediments, but are actually thriving long after burial.

32

Figure 1. (a) Location of samples taken during IODP X347 and used in this study. (b) Active
microbial community structures based on 16S rRNA transcript sequencing. Bacterial and
archaeal 16S rRNA transcripts were sequenced using domain-specific primers. Percentages
represent the relative abundance of each phylum with respect to the total within each domain.
The sample names are indicated below the Archaea.
33

Figure 2. Heatmap showing log transformed relative expression of genes related to cell division,
DNA replication and repair, and chemotaxis. Darker colors indicate higher numbers of reads
assigned to that gene.

34
Figure 3. Expression of key metabolic genes mapped onto biochemical pathways. Hashed boxes
represent genes where some but not all necessary subunits were expressed. Asterisks in boxes
represent samples with the highest percent of reads assigned. NP-P is nitrophenyl phosphate, NP
is nitrophenyl. See Supplementary Table 3 for a detailed list of gene subunits, abbreviations, and
number of assigned transcripts.

Figure S1. Assignment of transcripts to KEGG categories as a percent of all sequences per
sample.

35
“We may brave human laws, but we cannot resist natural ones.”
– Jules Verne, 20,000 Leagues Under the Sea

36
3. Climate influences microbial organic matter degradation in Baltic Sea
sediments
In collaboration with Clemens Glombitza, Jordan T. Bird, Hans Røy, Bo Barker Jørgensen,
Karen G. Lloyd, Jan P. Amend, and Brandi Kiel Reese.
Abstract
Globally, marine sediments are a vast repository of organic matter which is degraded through
various microbial pathways, including polymer hydrolysis and monomer fermentation. The
sources, abundances, and quality (i.e., labile or recalcitrant) of the organic matter and the
composition of the microbial assemblages vary between sediments. Here, we examine sediment
microbial communities from the Baltic Sea and the nearby Kattegat to determine connections
between geochemistry and the community potential to degrade organic carbon. Baltic Sea
sediments are an archive of extensive climate shifts over the last glacial-interglacial cycle.
Metagenomic sequencing of sediments from depths of 0.25 to 67 meters below seafloor having a
variety of depositional conditions revealed that diverse organic matter hydrolysis genes were
present and were in higher relative abundances in those sediments that contained more organic
matter. Metatranscriptomic analysis of two samples demonstrated that many of these genes were
transcribed. Some of the variation in deduced pathways correlated to carbon content and
depositional condition. Fermentation-related genes were found in all samples, and encoded for
multiple fermentation strategies. Notably, alcohol production genes were amongst the most
abundant genes. These results indicate that fermentation is not only widespread, but includes
multiple pathways, such as ethanol and acetate fermentation. Heterotrophic gene abundance
reflected depositional history and current in situ conditions. This study is a step towards a more
37
complete understanding of microbial food webs and the impacts of climate variation on sediment
communities.
Introduction
Organic matter (OM) burial in marine sediments sequesters carbon over geologic time
and regulates climate. Globally, marine sediments store 7.8×10
22
g of carbon, including as
organic matter that comes from both terrestrial and marine sources (Mackenzie et al., 2004).
Marine OM is generally more nitrogen rich than terrestrial OM. It contains carbohydrates and
proteins derived largely from water column organisms, compared with carbohydrates, such as
cellulose and lignin, derived from vascular plants in the terrestrial component (1, 2). The
contributions of these distinct organic pools to marine sediment varies between locations,
climates, and geologic times (e.g., (Calvert, 1987; R. Stein, 1990; Aller and Blair, 2004; Tao et
al., 2016). How these sources impact sedimentary carbon cycling and resident organisms is an
area of active research.
The marine sedimentary biosphere holds an estimated 5 x 10
29
prokaryotes (Parkes et al.,
2014), which some of which live in deeply buried sediments (Schippers et al.) and exert
significant control on biogeochemical cycles (e.g., (D’Hondt, Spivack, Pockalny, Ferdelman,
Fischer, Kallmeyer, Abrams, Smith, Graham, Hasiuk, Schrum, and Stancin, 2009b). Sediment
microbe metabolism varies between sediment settings, depending partially on nutrient, electron
acceptor, and electron donor availabilities (Röling et al., 2002; Lloyd, Lapham, and Teske,
2006b; Flores et al., 2011). Surface microbial communities, temperature, recalcitrance of
sediment OM, and depositional conditions also influence the composition and activities of the
sedimentary biosphere (Inagaki, Nunoura, Nakagawa, Teske, Lever, Lauer, Suzuki, Takai,
Delwiche, Colwell, Nealson, Horikoshi, D'Hondt, et al., 2006; Finke and Jorgensen, 2008;
38
Hamdan et al., 2012; Canion et al., 2014; Starnawski et al., 2017). In near-shore environments,
such as inland seas and along continental margins, organic loading to the sediment drives the
development of the microbial ‘anaerobic food web’ (Middelburg et al., 1993; Jorgensen, 2006)
and references therein). In organic-rich sediments, microbes in the upper few meters of sediment
below the seafloor use electron acceptors (e.g., O
2
, NO
3
-
, Mn(IV), Fe(III), SO
4
2-
) in order of
declining energy yields from organic matter respiration, ending with methanogenesis as the
dominant process (Froelich et al., 1979). Throughout the sediment column, heterotrophic
metabolisms are critical to breaking down the complex macromolecules and producing smaller
organic compounds, which feed into respiration and methanogenesis (Alperin et al., 1994).
However, it is difficult to distinguish between heterotrophic pathways in situ due to the large
range of bioavailable organic substrates, the micromolar concentrations of substrates and
products, the diversity of active microbial populations, and the number of pathways involved in
OM remineralization in sediments (Wellsbury and Parkes, 1995; Nealson, 1997; Arnosti, 2011;
Arnosti et al., 2014). Recently, advances in molecular biology (e.g., metagenomics and
metatranscriptomics), enzymatic assays, and organic geochemistry have allowed for more exact
studies of environmental OM degradation by microorganisms (Glombitza et al., 2014; Berlemont
and Martiny, 2016; Steen et al., 2016).
The Baltic Sea is an ideal location to study microbial organic matter mineralization. The
Baltic Sea is a shallow intracontinental sea which receives terrestrial inputs from rivers and
runoff and marine inputs from the North Sea via the Skagerrak and Kattegat (Andersson et al.,
1992). This has created a salinity gradient both laterally into the Baltic Sea and vertically in the
water column (Kullenberg and Jacobsen, 1981). Regional anoxia is frequent in the deepest basins
(Conley et al., 2009), and the sediments are rich in organic matter due to eutrophication and to
39
high sedimentation rates (up to 500 cm per 1000 years). Towards the end of the last glaciation,
the melting of the Scandinavian Ice Sheet caused dramatic environmental changes.
Approximately 16,000 years ago, as the basin was still partially covered by the ice sheet, the
glacial Baltic Ice Lake started to form (Houmark-Nielsen and Henrik Kjær, 2003). Between the
start of the Holocene 11,700 yr BP (years before present) to approximately 10,700 BP, a
connection of the ice lake to the North Sea caused a brief brackish phase of the basin, the Yoldia
Sea (T. Andrén, Björck, E. Andrén, Conley, Zillén, and Anjar, 2011a). This was followed by the
low primary-productivity, freshwater Ancylus Lake phase. Sediments deposited in both the
Baltic Ice Lake and the Ancylus Lake are organic-poor with relatively low pore water salinities
relative to brackish phases (T. Andrén, Björck, E. Andrén, Conley, Zillén, and Anjar, 2011b). By
circa 9,800 BP, a gateway from the Baltic Sea to the North Sea was established, and the entire
basin became a brackish-marine sea with high productivity (Sternbeck et al., 2000) from the
onset of the Littorina Sea phase. Sediments deposited in the Littorina Sea and in the modern
Baltic Sea are organic-matter rich, highly reducing, and methanogenic (T. Andrén, Barker
Jørgensen, et al., 2015; Egger et al., 2017). Overall, the contrasts in depositional conditions
within the past ~16,000 years (fresh versus saline/brackish, organic-rich versus organic-poor,
etc.) create natural gradients that may influence the types and pathways of organic matter
mineralization possible in the present-day microbial communities.
Here, we investigate sediments from several depths at four sites that differ in organic
matter content, depositional histories, and salinities: three sites within the Baltic Sea Basin and
one site in the Kattegat, the basin’s marine connection to the North Sea. DNA and RNA
sequencing (i.e., metagenomics and metatranscriptomics) were employed to determine the
organic matter mineralization pathways present in the sediments, assess which pathways were
40
likely active, and connect these metabolisms to in situ geochemistry and past depositional
conditions.
Methods
Sample collection
Samples M59E-15m, M59E-67m, M60B-24m, M60B-37m, M63E-11m, M63E-47m,
M65C-3m, M65C-10m, and M65C-30m were collected by the Integrated Ocean Drilling
Program (IODP) aboard the Greatship Manisha in September-November, 2013. The sediment
cores were taken by advanced piston coring. On board, cores were cut into 1.5 m sections,
sampled for PFC contamination (to assess the amount of drilling disturbance), scanned with a
fast-track Multiple Scanning Core Logger, and sectioned into whole round cores in a 12°C
microbiological container on board the ship. Sediment cores for nucleic acid analyses were
immediately frozen at -80°C on ship and shipped to land-based laboratories on dry ice (cf. (T.
Andrén, Jorgensen, et al., 2015).
To sample the top few meters of sediment, which were not recovered in the drilling
process, two cruises lead by the Center for Geomicrobiology at Aarhus University on the R/V
Aurora were undertaken. In September, 2014, site M59 was revisited. Sample M59E-0.25cm was
collected through gravity coring, followed by subsampling through windows cut into the
sediment core liner, and insertion of sterile 20 ml syringes with the ends cut off. In June, 2016,
site M65C was revisited, and sample M65C-0.25m was collected through Rumohr coring (as in
(Beulig et al., 2018). Subsampling was similar to sample M59E-0.25m, in which sterile cut-off
20 ml syringes were used to subsample the 25cm depth horizon. Care was taken to avoid
potential seawater contamination by visually inspecting the cores for seawater intrusion and
41
extracting DNA from sediment taken from the interior of the core. Both samples were
immediately frozen at -80°C on ship, and were shipped to the United States on dry ice.
Geochemical Analysis

Pore water samples for volatile fatty acids (VFA) analysis were retrieved with Rhizon
soil-moisture samplers (Rhizosphere Research Products, Wageningen, The Netherlands)
(Seeberg-Elverfeldt et al., 2005) or were obtained by a hydraulic press (Manheim, 1966)
according to IODP standard protocols if the sediment was to compacted. Rhizons were pre-
cleaned with 50 mL Milli-Q® water (Ultrapure, Type 1) and stored in vacuum-sealed gas tight
bags (Glombitza et al., 2014). The samples were stored at -80°C in 4 mL borosilicate glass vials
(Zinsser Analytic, Germany) that were previously baked for 5 h at 450°C. Prior to the analysis,
the samples were defrosted and filtered through disposable Acrodisc® 13 mm IC syringe filters
(pore size 0.2 µm) that were rinsed with 10 mL Milli-Q (Ultrapure Type I) water directly before
use. The first 0.5 mL of pore water after filtration was discarded and a second 0.5 mL was used
for analysis. VFA concentrations, including formate, acetate, butyrate, and propionate, were
measured by two-dimensional ion chromatography mass spectrometry (2D IC-MS) as described
in detail in Glombitza et al. (2014). Briefly, in this method the first IC dimension is used to
separate inorganic ions, such as chloride, from VFAs. VFAs are trapped on a concentrator
column and subsequently separated in the second IC dimension. Quantification is achieved by
the mass spectrometer in the single ion-monitoring (SIM) mode. Detection limits are 0.37 µM
for formate, 0.19 µM for acetate, 0.12 µM for propionate and 0.09 µM for butyrate.
Quantification was achieved by a 3-point calibration with external standards of a mixture of
VFAs (formate, acetate, propionate) at different concentrations (i.e., 200, 500 and 800 µg L
-1
) in
IAPSO seawater standard (OSIL, UK).
42
Other geochemical data for IODP cores (all samples here except M59E-0.25m and
M65C-0.25m) were collected and analyzed as described in (T. Andrén, Jorgensen, et al., 2015).
Briefly, Rhizon soil-moisture samplers and core squeezers were used to retrieve sediment pore
water. Sulfate and Cl
-
were measured via ion chromatorgraphy using a Metrohm 882 compact
ion chromatograph (Metrohm AG, Herisau, Switzerland) at the University of Bremen. Methane
samples were collected from fresh core material, extruded into 8 mL of 1 M NaOH filled glass
vials, shaken and equilibrated, and measured on an Agilent A7890 gas chromatograph (Agilent
Technologies, Santa Clara, CA, USA) (T. Andrén, Jorgensen, et al., 2015; Egger et al.). TC and
TOC were sampled from 10 cm
3
of freeze-dried and ground sediment. TC measurements were
derived from approximately 65 mg of sample that was combusted, and evolved CO
2
was
measured on a LECO CS-300 carbon-sulfur analyzer (LECO Corporation, Saint Joseph, MI,
USA). TOC was measured from 65 mg of 12.5% HCl decalcified sediment, which was then
heated and evolved CO
2
was measured as above (T. Andrén, Jorgensen, et al., 2015).
For samples M59E-0.25 cm and M65C-0.25 cm, geochemical analyses were performed
as described previously (Beulig et al., 2018). Rhizon samplers extracted sediment pore water,
which was then acidified and measured at the University of Aarhus (Røy et al., 2014). Methane
samples were taken immediately after core retrieval, transferred into vials with 4 mL of saturated
NaCl, capped, stored at -20°C, and measured on an SRI 310C gas chromatograph equipped with
a flame ionization detector SRI 310C (SRI Instruments, Torrance, CA, USA) (Beulig et al.,
2018).

43
Extraction
DNA from samples M59E-67m, M60B-24m, M60B-37m, M63E-11m, M63E-47m,
M65C-3m, and M65C-30m were extracted and sequenced as described in Marshall et al., 2017.
DNA from samples M59E-0.25m, M59E-15m, M65C-0.25m, M65C-3m, and M65C-10m was
extracted from sediment using the MoBio DNeasy Power Max soil kit (MoBio Laboratories,
Carlsbad, CA). Frozen sediment was chipped from whole round cores (M59E-15m, M65C-3m,
M65C-10m) or cut-off syringes (M59E-0.25m and M65C-0.25m) in a dedicated clean room at
Texas A&M University Corpus Christi. All instruments used were ethanol and flame-sterilized
and edges of the sediment core were avoided, and researchers wore face masks and hair nets to
avoid sample contamination. Between 5 and 10 g of sediment was extracted for each sample.
The manufacturer’s protocol was followed, including the final concentration step and
resuspension of DNA in 100 µl of molecular biology grade water. Sample free negative controls
(kit blanks) were processed and sequenced alongside the samples. These negative controls were
below detection limit (0.5 ng DNA ul
-1
) when measured using the Qubit DS high sensitivity kit,
and did not amplify when subject to PCR of the 16S rRNA gene and visualized on a 1% agarose
gel.
RNA was extracted as described in Zinke et al. (Zinke et al., 2017). Briefly, RNA was
extracted from sediment using the MoBio PowerSoil RNA kit (MoBio Laboratories, Carlsbad,
CA) following manufacturer’s instructions. Approximately 7.5 g of sediment from each sample
was extracted, which was divided over three reactions (2.5 g each). The final RNA-containing
pellets were sequentially combined in 50µl of PCR-grade RNase-free water. No-template
contamination controls were run alongside the samples and were assessed for sterility through
PCR amplification and agarose gel visualization. Total RNA extractions were treated with
44
Ambion Turbo DNase (ThermoFisher Scientific, Waltham, MA) according to manufacturer
protocols. Resulting RNA purity and quantity was checked using the Eppendorf Biospectrometer
(Eppendorf, Hauppauge, NY) and by reverse transcription of the 16S rRNA gene transcript
followed by PCR amplification of the complimentary DNA (cDNA).

Sequencing
Metagenomes from samples M59E-0.25m, M59E-15m, and M65C-0.25m were
sequenced at the Marine Biological Laboratories (Woods Hole, MA, USA). Metagenomic library
preparation and sequencing followed the Census of Deep Life protocol as described in Vineis et
al. (Vineis et al., 2016). The modifications to the Vineis protocol were that the sequencing
platform was the NextSeq (Illumina, San Diego, CA, USA), which produced 150 base pair (bp)
long paired end reads, and no microbiome enrichment step was conducted.
Metagenomes from samples M65C-3mbsf and M65C-10mbsf were sequenced at the
Research and Testing Labs (RTL; Lubbock, TX, USA). Libraries were prepared using the Kapa
HyperPlus kit (Kapa Biosystems, Wilmington, MA, USA) according to manufacturer
instructions with the following modifications: ligation was increased to 1 hour at room
temperature, post-ligation cleanup used 0.75X beads, and the post-amplification cleanup bead
concentration was increased to 0.7X. Libraries were sequenced on the Illumina HiSeq 2500
(Illumina, San Diego, CA, USA), producing 150 bp long paired end reads. These data can be
found in the National Center for Biotechnology Information (NCBI) Sequence Read Archive
(SRA) under project SRP142652.
Metagenomes from all other samples here were sequenced as described in Marshall et al.
(2017), and can be found in the NCBI SRA under project SRP068645. Briefly, DNA was
extracted using the FastDNA SPIN kit for Soil DNA extraction (MP Biomedicals, Santa Ana,
45
CA, USA). Resulting DNA underwent library preparation with the Nextera XT DNA Library
Preparation kit (Illumina, San Diego, CA, USA), and were sequenced on the Illumina HiSeq
2000 platform to produce 100 bp paired end reads.
Metatranscriptomes were sequenced as described in Zinke et al. (Zinke et al., 2017), For
metatranscriptomes, complementary DNA synthesis was performed using the QuantiTech Whole
Transcriptome Amplification kit following manufacturer’s instructions (Qiagen, Hilden,
Germany). Libraries were generated from cDNA using the Nextera DNA sample preparation kit
(Illumina, San Diego, CA) and were sequenced on the Illumina HiSeq 2500 platform (Illumina,
San Diego, CA) for 500 cycles with 250 bp paired-end chemistry. These data can be found under
NCBI project SRP108285. See Supplemental Table 1 for sequencing statistics.

Bioinformatics
Quality control and assembly
Reads were trimmed using the program Trim Galore! v0.4.3 (Babraham Bioinformatics,
Cambridge, UK) in paired-end read mode, with a minimum quality score of 25, a maximum 4
low quality bases the read was trimmed, and the read length must be a minimum of 80 bp long
post trimming. Samples were dedeplicated using Super Deduper v2.0 with default settings
(starting location of 10 bp, 25 base pairs in the unique ID) (K. R. Petersen et al., 2015). All
metagenomes were co-assembled using Megahit v1.0.3-29-g707d683 with a minimum contig
size of 1000 bp (Li et al., 2015; 2016). Default kmer sizes of 21, 29, 39, 59, 79, 99, 119, and 141
were used for assembly. Contig names were simplified using anvi-script-reformat-fasta in anvi’o
v2.4.0 (Eren et al., 2015).

46
Read mapping and profile generation
Metagenomic and metatranscriptomic reads were mapped to the assembled contigs using
Bowtie2 v2.2.5 using the ‘sensitive’ end-to-end setting (Langmead and Salzberg, 2012).
Resulting sam files were converted to bam files using samtools v1.5, and these files were
converted to anvi’o-compatible bam files in anvi’o. An anvi’o database was created from the
contigs, which included Open Reading Frame (ORF) determination using Prodigal (Hyatt et al.,
2010). Each sample was profiled against the contigs database using the anvi’o command anvi-
profile. A full project database was constructed from these profiles, and included information
about gene coverage and detection (percent over which the ORF was mapped by reads). Tables
with gene coverage by sample and gene detection by sample were exported using the anvi’o
command anvi-export-gene-coverage-and-detection.
Reads that remained unmapped were merged using FLASH v1.2.11 (Magoč and
Salzberg, 2011) with a minimum overlap of 10 base pairs and a maximum overlap of 100 base
pairs.
Taxonomic and function assignments of genes
Protein coding sequences were exported from the contigs for taxonomic and functional
assignment using the anvi’o command anvi-get-aa-sequences-for-gene-calls. These sequences
were compared to the NCBI non-redundant database (accessed December, 2016) using the blastp
mode of DIAMOND v0.8.36 (Buchfink et al., 2014) with ‘sensitive’ setting and allowing only
one match per sequence allowed. DIAMOND results were uploaded to MEGAN v6.10.2 and
taxonomy was assigned using the weighted Lowest Common Ancestor (LCA) assignment
algorithm (Huson et al., 2011; Huson, Beier, Flade, Górska, El-Hadidi, Mitra, Ruscheweyh, and
Tappu, 2016b) with a minimum support percent identity of 0.3 (i.e., a taxon must have at least
47
0.3 percent, or 2,888 ORFs, assigned to be considered a ‘real’ hit). Taxonomic assignments for
ORFs were exported as a tab separated file and parsed in R.
Function was assigned to assembled genes using INTERPROSCAN v5.26-65.0 (Jones et
al., 2014) against the Pfam v31.0 (Finn et al., 2013) and TIGRFAM v15.0 databases (Haft et al.,
2012) (accessed September, 2017) using the precalculated lookup service. Functional
assignments were exported as a tab separated file and parsed in R. Genes annotated as peptidase
or CAZyme coding were screened for export peptides using signalp targeted for bacteria (gram
positive and negative) and eukaryotes (T. N. Petersen et al.), and were screened using psortb
targeted for archaea (Yu et al., 2010). PFAM functions were manually searched in the PFAM
database in August - November, 2017. Pathway reconstruction was based on pathways in the
KEGG (Kanehisa et al., 2016) and Metacyc databases (Caspi et al., 2013) and published
literature.
Coverage of coassembled genes was parsed from coverage profiles produced in anvi’o.
These were parsed by determining which genes contained more than 50% gene detection, and
averaging gene coverages by PFAM assignment. Abundances of reads were calculated by
dividing the number of reads per metagenomes that mapped to a PFAM or TIGRFAM
assignment by the total number of reads in the metagenome. Resulting percentages were
multiplied by one million to give a relative abundance per million reads. One limitation of using
relative percentages in metagenomic analysis is that it cannot be determined if the absolute
abundance of a set of genes is the same in the environment, and if the number of other functions
or microbes are different, resulting in perceived differences in the environment. Conversely, the
number of genes for a process may be different, but normalization causes these values to appear
similar.
48
Reads that did not map to the assembly were examined for function by comparing to the
nr database as above using DIAMOND in sensitive blastx mode, with only one match allowed
per read. Results were imported into MEGAN and were compared to the Interpro2Go database.
These results were inspected for gene assignments not present in the coassembly, and results
used in determining potential methane cycling processes. These reads were not assigned
taxonomy, however, and were potentially derived from non-methanogenic or methanotrophic
organisms, so caution should be taken in assuming some phylogenetically widely-distributed
genes are from methane cycling organisms.
Statistical analyses
All statistics, including linear modeling, canonical correspondence analyses, t-tests, and
permutational analysis of variance (PERMANOVA), were performed in R version 3.4.2.
Ecological tests were performed using the vegan package version 2.4-4 (Dixon, 2003). The
ggplot2 package (Wickham, 2011) was used to create graphics.
Results
Site description
Sediment samples were collected from three locations in the Baltic Sea Basin (sites M59,
M63, and M65) and one location in the Kattegat Strait (site M60) (Fig. 1a). The water depths
were between 31.2 m and 437.1 m, with cored sediment depths from 0.25 m to 67 m below
seafloor (mbsf) (Table 1). Baltic Sea Basin samples were deposited either during non-glacial
conditions in the Holocene epoch (samples M59E-0.25m, M59E-15m, M63E-12m, M65C-
0.25m, M65C-3m, and M65C-10m) or during glacial conditions during the upper Pleistocene or
lower Holocene epoch (samples M59E-67m, M63E-47m, and M65C-30m) (T. Andrén, Barker
Jørgensen, et al., 2015). The samples from 60B were deposited in the upper Pleistocene after
49
deglaciation of the Kattegat, which provided seawater influx from the North Sea (Friberg). The
present-day salinity of these samples varied, with the freshest pore water salinities found in the
deepest basin samples (sample M59E-67m: 7.48, sample M63E-47m: 1.67, sample M65C-30m:
2.78) (T. Andrén, Jorgensen, et al., 2015). The salinity variations have been attributed to glacial
depositional conditions during which basin waters were heavily influenced by melting glacial
water (Fig. 1b) (T. Andrén, Barker Jørgensen, et al., 2015). Total Carbon (TC) content of
sediment in our samples varied between 0.70% and 5.95% dry weight (d.wt.), with the largest
value measured in sample M59E-15m (Table 1). In this sample, most of the TC was present as
Total Organic Carbon (TOC), which was 4.97% d.wt. (T. Andrén, Jorgensen, et al., 2015). The
lowest levels of TC (0.70% d.wt.) and TOC ( ≥ 0.48% d.wt.) were in samples deposited under
glacial conditions (Table 1, Fig. 1c). Methane was present in multiple samples, with the highest
measured concentration at 10.10 mM (Table 1). Because the cores experienced significant
degassing during sampling, the reported values are minimum in situ methane concentrations
(Andrén et al, 2015; Egger et al., 2017). Samples within the basin originated from below the
main sulfate zone and measured sulfate concentrations were therefore generally low (0.01 - 0.22
mM). The exception is site M60 samples which contained concentrations ≥ 6.59 mM (T.
Andrén, Jorgensen, et al., 2015). Generally, greater TOC and methane concentrations were
observed in the non-glacial samples with high pore water salinities.
Sequencing and assembly
Coassembly of all metagenomes from every sample produced 557,851 contigs of length ≥ 1,000
base pairs (bp). The coassembly contained a total of 1.07 gigabases (Gb), with a maximum
contig length of 159,111 bp and an average length of 1,923 bp (Supplemental Information Table
1). Between 15.43% and 35.95% of metagenomic and metatranscriptomic reads in each sample
50
mapped to the assembly. From the coassembly, 1,477,923 protein coding sequences were
predicted. When compared to all proteins in the Interproscan v66.0, PFAM v31.0, and
TIGRFAM v15.0 databases (accessed October 2017), 1,074,069 predicted genes were annotated
to be within functional families or assigned to putative/hypothetical protein families in at least
one of these databases. The remaining protein coding sequences did not correspond to known or
hypothetical genes.
Protein utilization
Peptidase encoding genes with export peptides (e.g., extracellular peptidases) were
examined to predict extracellular enzymatic protein degradation, and were assigned to PFAM
peptidase families and described here using the MEROPS peptidase nomenclature (Rawlings and
Barrett, 1993). Peptidase families M24 (methionine aminopeptidase), S8 (subtilase), and M20
(glutamate carboxypeptidase) were the relatively most abundant peptidase-encoding genes, with
genes for families C25 (gingipain) and M48 (Ste24 endopeptidase) also relatively abundant. Of
these genes, three (glutamate carboxypeptidase, gingipain, and Ste24 endopeptidase) were
significantly positively correlated with TOC (linear regression, p < 0.05) and methionine
aminopeptidase was significantly related to both TOC and salinity. Subtilase gene abundance
was not significantly associated with salinity or TOC, but was weakly significantly higher in
marine than lacustrine depositional conditions (t-test, p = 0.049) (Supplemental Fig. 1). All
metagenomic reads mapping to the extracellular peptidase encoding genes were significantly
more abundant in samples with the greatest TC (PERMANOVA, p = 0.007). Relative abundance
also significantly positively correlated with salinity (p = 0.023), marine depositional conditions
(p = 0.033), TOC (p = 0.039), and formate concentrations (p = 0.021). Transcribed peptidases
included alkaline D-peptidase (165 normalized reads per million (nrpm) in 59E-15m and 31
51
nrpm in 63E-12m), methionine aminopeptidase M24 (194 nrpm in M59E-15m and 156 nrpm in
63E-12m), gingipain (168 nrpm in M59E-15m and 13 nrpm in M63E-12m), and subtilase (153
nrpm in M59E-15m and 282 nrpm in M63E-12m).
Carbohydrate utilization
Genes that potentially mediate complex carbohydrate degradation (Carbohydrate Active
enzymes, or CAZymes) and contained cellular export signals were examined. Glycoside
Hydrolases (GH) are critical proteins in hydrolyzing complex carbohydrates. Genes were found
here for GH families with fucosidase, amylase, lysozyme, chitinase, cellulase, and xylanase
activities (Fig. 3). Collectively, these GHs can degrade carbohydrates from various sources,
including plants and algae (Lombard et al., 2013). The most abundant exported GH families
included families 5 (cellulase, 0-15 nrpm across all samples), 10 (xylanase, 1-21 nrpm), 23-25
(lysozymes, 4-21 nrpm), and 29 (fucosidase, 0-38 nrpm) (Fig. 3). All of these GHs except the
lysozymes showed significant positive relationships with TOC (linear regression, p < 0.05)
(Supplemental Fig. 2). The lysozyme relative abundances significantly correlated with more
marine Holocene versus glacially influenced depositional conditions (t-test, p = 0.0002).
Gene annotations were searched for potentially exported Carbohydrate Binding Modules
(CBM), which are associated with GHs and bind target substrates (Shoseyov et al., 2006). The
most abundant CBM genes found were mostly within families binding cellulose (family 10, 0-4
nrpm), chitin (family 5/12, 0-15 nrpm), and cell wall material, such as peptidoglycan (family 50,
4-31 nprm) (Notenboom et al., 2001; Boraston et al., 2004; Machovič Martin and Štefan, 2008)).
Linear modeling showed that chitin-targeting CBMs were significantly correlated with TOC (p =
0.0002 for both). Other putatively exported CAZymes found included pectate lyases and alginate
lyases, though alginate lyases were in low relative abundances (Fig. 3).
52
Permutation multivariate analysis of variance tests showed that total exported CAZyme
abundances corresponded most strongly with TOC (p = 0.009). These CAZymes showed weaker
but still significant correlations to salinity (p = 0.02), marine vs lacustrine (including glacial lake)
depositional environment (p = 0.023), and approximate age of the sediments (p = 0.01).
The two metatranscriptomes showed that both GHs and CBMs were transcribed in situ.
Transcripts mapping to cellulases, fucosidases, invertases, lysozymes, plant matter (other than
cellulose) targeting GHs, and oligosaccharide-targeting GHs were relatively abundant, which
was similar to the most abundant GHs in the metagenomes. Of the transcribed CBMs, chitin and
peptidoglycan-targeting CBMs were the relatively most abundant, with some transcripts
mapping also to starch and plant matter binding CBM encoding genes.
Microorganisms can hydrolyze complex carbohydrates into smaller molecules, which are
then respired and further fermented. Genes related to these processes were found in all samples
(Supplemental Fig 2). The most relatively abundant genes were assigned to the FGGY family (1
- 143 nrpm), which is a broad family of carbohydrate kinases such as gluconokinase,
xylulokinase, fuculokinase, ribulokinase, and rhamnulokinase (Ying Zhang et al., 2011). Fucose
transporter genes were found in all non-glacial basin metagenomes and were expressed in the
metatranscriptomes, but no reads mapped to these genes in the glacial samples. Deacetylases
were also found in the metagenomes and expressed in the metatranscriptomes. These included
diacetylchitobiose deacetylase, which contributes to chitin degradation by removing acetyl
groups from diacetylchitobiose (average of 34 nrpm in non-glacial Baltic sediments and 8 nrpm
in glacial samples) (Tanaka et al., 2004).

53
Fermentation
The metagenomes and metatranscriptomes were analyzed for genes related to
fermentation. Genes related to pyruvate conversion to acetate were abundant, including both
pyruvate ferredoxin oxidoreductase and pyruvate formate lyase, and these genes were transcribed
in the two metatranscriptomic samples (Fig. 4). Pyruvate ferredoxin oxidoreductase genes (36-
147 nrpm) were more relatively abundant than pyruvate formate lyase genes (0-99 nrpm) in all
metagenomes. Linear regressions showed pyruvate ferredoxin oxidoreductase gene abundance
was significantly correlated to TC (p = 0.02) but not TOC, and pyruvate formate lyase was
significantly positively correlated with TOC (p = 0.0035) and formate concentration (p =
0.00003). In the metatranscriptomes, these two genes were similarly abundant, with pyruvate
formate lyase slightly more abundant in each sample (M59E-15m: 141 nrpm pyruvate ferredoxin
oxidoreductase, 231 nrpm for pyruvate formate lyase; M63E-12m: 72 and 77 nrpm). Acetyl-CoA
hydrolase/transferase family genes, which facilitate the production of acetate from acetyl-CoA or
the production of acetyl-CoA from acetate and acyl-CoA, and acetate kinases were found in all
samples. Acetate kinase genes were significantly correlated with TOC content (linear
regressions, p = 0.00025), formate concentrations (p = 0.00012), and with salinity (p =
0.0001632) (Supplemental Fig. 4).
Genes annotated as alcohol dehydrogenases, which can produce ethanol during
fermentation, were also present and relatively abundant in all samples. In non-glacial samples,
the relative abundance of alcohol dehydrogenases was greater than or approximately equal to the
relative abundance of pyruvate genes (Fig. 4). Most of the non-glacial samples show a
dominance of zinc-binding alcohol dehydrogenases relative to iron-containing alcohol
dehydrogenases. In contrast in the glacial samples, these genes were either in similar relative
54
abundance (samples 59E-67m and 63E-47m) or iron-binding dehydrogenases were dominant.
Zinc binding alcohol dehydrogenases were significantly correlated with TOC (linear modeling, p
= 0.001), but iron-binding dehydrogenases were not found to be significantly correlated with
geochemical parameters tested here, including TOC, fatty acid concentrations, and salinity. In
the metatranscriptomes, the two types of alcohol dehydrogenases had similar relative numbers of
assigned reads (Fig. 4).
Other present and expressed fermentation-related genes included 2-hydroxylglutaryl-CoA
dehydratase encoding genes. This enzyme catalyzes a key step in glutamate fermentation
(Schweiger et al., 1987; Hans et al., 1999). Glycerol oxidation genes dihydroxyacetone kinase
and glycerol dehydratase (Daniel et al., 1995) were found in low abundances in the
metagenomes, and at least one of these genes was expressed in each metatranscriptome. Genes
encoding acetoacetate decarboxylase, which produces acetone and CO
2
from acetoacetate during
fermentation (Amador-Noguez et al., 2011), were present in all Holocene metagenomes (Fig. 4),
and were strongly correlated to TOC (p = 0.0042). The overall abundance of fermentation genes
in samples was significantly related to TOC (MANOVA, p < 0.05), but not to salinity,
marine/lacustrine sediment deposition, any fatty acid examined here, and sediment age.
Methane
Methane is a critical end-product in organic matter mineralization, and can also act as
electron donor in anaerobic and aerobic methanotrophy. Assembled contigs and unmapped reads
were searched for genes related to methane metabolism. Complete or near complete pathways for
hydrogenotrophic and methylotrophic methanogenesis were found in all samples, except 60B-
37m. Methanol utilization genes were also found in many of these sediments, but were not found
in two of the glacial samples (M59E-67m and M65C-30m). Some genetic support for
55
acetoclastic methanogenesis was found in several samples, indicated by the presence of acetyl-
CoA synthetase and acetate kinase genes assigned within methanogenic orders (Fig. 5). Genes
for methane cycling were expressed in both metatranscriptomes, including mcr subunits and
methylamine utilization subunits.
Discussion
Organic matter degradation
Protein degradation
Protein-derived compounds represent over 20% of organic matter in some sediments
(Wakeham et al., 1997; Dauwe and Middelburg, 2003). In organic-rich sediments, the microbial
genetic potential to degrade macromolecules has previously been demonstrated through genomic
analyses. In our study, peptidase-encoding genes with export signal sequences were abundant in
marine Holocene samples, and in all samples were more abundant than predicted CAZymes.
Relative abundance of peptidases was significantly positively correlated with sediment TC and
TOC. Recently, Schmidt and Steen (2016) performed Baltic Sea sediment incubation studies
with labeled peptidase substrates, which showed that extracellular peptidases from site M59 were
active down to 55 meters below seafloor in the organic rich Holocene mud (Schmidt, 2016).
Archaeal genomes isolated from Bathyarchaeota and Thermoplasmatales from Aarhus Bay
sediment contained peptidases, including gingipain and clostripain, which also appeared to be
active in enzyme assays of whole sediment (Lloyd et al., 2013). It is possible that these
organisms catabolize extracellular peptidases for energy (60). Marshall et al. (2017) showed that
Calditrichiaeota, a bacterial phylum found in the Baltic Sea and in marine sediments globally,
also likely degrade extracellular proteins for energy (Marshall, Starnawski, et al., 2017). The
results herein agree that protein-degradation for energy acquisition is an important heterotrophic
56
strategy in the marine deep biosphere, and based on metatranscriptomics, was likely active in
these sediments.
Plant matter degradation
Near shore environments receive globally 0.4 Pg yr
-1
of terrestrial organic matter from
rivers, including plant matter, such as lignin, cellulose, and xylan (Hedges et al., 1997). Genes
for CAZymes associated with plant matter degradation were abundant in BSB sediments here,
including multiple glycoside hydrolases and carbohydrate binding modules. The abundances of
these genes were positively correlated with TOC content (Supplemental Figure 3). Many
CAZyme genes were transcribed in Baltic sediments at 12 and 15m below seafloor, suggesting
active plant matter degradation (Fig. 3). CAZyme-encoding genes have been found in estuarine
sediments (Seitz et al., 2015; B. J. Baker et al., 2015), in deep sea sediments (Klippel et al.,
2014), and in the top 10 cm of sediment from the Landsort Deep (Baltic Sea) (Thureborn et al.,
2016). Recently, genes for plant-derived OM degradation were found to be transcribed in Peru
Margin sediments (Orsi et al., 2018). Our results demonstrate that plant matter degradation
potential is present deep into sediments, and is active to depths of at least 15 mbsf.
Algal biomass degradation
Macroalgae are primary producers which can rapidly produce biomass in coastal
sediments (up to 3 kg C m
-2
yr
-1
) (Chung et al., 2011),and references therein), but relatively little
is known about algal biomass degradation in sediments (Matos et al., 2016). In sediments
examined here, alginate lyase encoding genes were found in low but detectable abundances at
sites M59E, M63E, and M65C in sediments deposited during the Holocene, and were expressed
in M63E-12 mbsf and 59E-15 mbsf. Fucoidan is produced by brown algae and seaweeds ((Ale et
al., 2011), and references therein). Genes encoding fucosidases were abundant in sediments
57
deposited during marine-brackish periods, and expressed within the metatranscriptomes.
Recently, Matos et al. (2017), showed that the potential to degrade complex, algae-derived
organics was present in diverse microorganisms in the top 5 cm of sediment in the Baltic Sea,
near Svalbard in the Barents Sea, near Tierra del Fuego at the southern tip of South America, and
off the coast of Antarctica (Matos et al., 2016). Our study extended this by detecting alginate
lyase encoding transcripts, which is indicative that microbes could be degrading algae-derived
organics throughout the Baltic Sea thousands of years after deposition.
Chitin degradation
Chitin is the second most produced biomolecule worldwide (Gooday, 1990), and is
abundant in marine systems. Here, we found that putative chitinase encoding genes occur in
metagenomes within all of the Holocene sediments in lower relative abundances than genes
encoding for most other carbohydrate degrading processes. Studies of other nearshore marine
systems have found chitinases in surficial sediments (Boyer, 1994; Wu et al., 2008; Teplyuk et
al., 2017). Potentially, the relative scarcity of chitinase genes and transcripts here is related to the
relatively fast degradation of chitin in sediments (in under two years post-deposition in some
cases) (Poulicek and Jeuniaux, 1991). Our sediments range from hundreds to thousands of years
old, so it is likely that most of the bioavailable chitin has been degraded, and that chitin is not a
significant source of nutrient in deeper sediments. In summary, heterotrophy in the organic-rich
Holocene sediments of the Baltic is diverse, persists long after sediment deposition, and is
supported by both terrestrial and marine organic matter.

58
Fermentation is active and ubiquitous
We found genes for VFA production and consumption throughout the sediment column
and across sites. We also found that acetate and ethanol producing genes were the most dominant
expressed fermentation genes in both of our metatranscriptomes, indicating that these processes
are active in at least some of the sediments here. Acetate was found in concentrations of up to 37
µM in porewaters in these samples, confirming acetate production, but no statistically significant
correlation between gene relative abundance and acetate concentration was found. This is not
surprising, since VFA pore water concentrations are under the control of VFA-consuming
organisms as opposed to fermenters (Glombitza et al., 2015; Beulig et al., 2018). However,
several fermentation and acetate related genes, including acetate kinase and pyruvate formate
lyase (which produces formate) were significantly correlated to formate concentrations, which
ranged from below detection limit to 10.4 µM in our samples. Formate was not correlated to any
geochemical conditions here, but acetate kinase and pyruvate formate lyase both were strongly
associated with TOC. Potentially, TOC influences fermentation gene abundances, which then
influence the concentration of formate, though the precise controls on formate concentrations
remain unknown (Glombitza et al., 2015).
Previous studies of fermentation in marine sediments have found multiple fermentation
pathways possible, including those which produce acetate, propionate, and ethanol (Kirchman et
al., 2014). Kirchman et al. compared the relative abundances of these pathways in metagenomes
from methane-rich marine sediments in the Arctic (Kirchman et al., 2014), Nyegga cold seeps
near the Norwegian margin (Stokke et al., 2012), and Tonya cold seeps near California
(Håvelsrud et al., 2011). They found that both acetate and ethanol fermentation pathways were
abundant in all sediments, which was similar to the results here. While relative gene or transcript
59
abundance does not predict the in situ turnover rates, these findings indicate that the types of
fermentative pathways present vary among sediments, and likely are influenced by the
concentration and types of sediment TOC.
Fermentation feeds microbial respiration (such as sulfate reduction) and methane
production, and our data point to multiple methanogenic pathways. These include
hydrogenotrophic methanogenesis and methane production from methanol and methylated
compounds. Interestingly, based on radiocarbon labeled acetate amendments, Beulig et al. found
acetate-derived methane in Bornholm Basin (site M65) sediment was not produced through
acetoclastic methanogenesis (Beulig et al., 2018). They concluded that an unknown microbial
partner oxidized acetate to CO
2
, which was then reduced to CH
4
by methanogens. The traditional
acetoclastic methanogenesis pathway in which acetate kinase and phosphotransacetylase or
acetyl-CoA synthetase produce acetyl-CoA, which then enters the methanogenesis pathway by
the action of CO dehydrogenase/acetyl-CoA synthase. Acetate kinase, phosphotransacetylase,
and acetyl-CoA synthetase genes that could be assigned to known methanogenic taxa that were
in present in exceedingly low abundances (less than or equal to 1.2 nrpm in several sediments
here (including M59E-0.25m, M59E-15m, andM 65C-0.25m), and were not found at all in most.
This agrees with Beulig et al (2018), in which direct acetoclastic methanogenesis could be
detected only at very low rates in sediments at site M65C where hydrogenotrophic
methanogenesis dominated (Beulig et al., 2018).
Understanding fermentation processes, which feed into other respiratory pathways, is
critical in understanding anaerobic organic matter degradation in sediments. Several
fermentation-related genes were significantly correlated with environmental parameters,
including salinity and carbon/organic matter content of the sediment, but many genes did not
60
correlate with the parameters tested here. Acetate remained a key fermentation product, as found
in other studies, but ethanol fermentation deserves further investigation in marine sediments.
Future characterization of marine sediment microbiomes, studies of VFA dynamics, and rate
measurement experiments are needed to illuminate controls on the types and amounts of
fermentation present and active.
Impacts of depositional condition on microbial community function
The assemblages of microbial communities in the subseafloor depend on the initial
microbial inoculum and sediment composition, and are selected for during burial (Walsh et al.,
2016; Starnawski et al., 2017). Past climatic events can influence deep seafloor community
compositions and functions long after burial (Inagaki et al., 2003; 2015; Orsi et al., 2017). The
deglaciation of the Baltic Sea Basin during the late Pleistocene and fluctuating water column
conditions throughout the Holocene is reflected in the lithology and in situ geochemistry of
Baltic Sea sediment (T. Andrén, Barker Jørgensen, et al., 2015). This makes the Baltic Sea an
ideal setting to study microbial community shifts along a climate archive. Previously, significant
differences have been observed in the microbial community structure based on deposition,
including halogenated compound degradation and C1 metabolisms, such as methane usage and
the Wood-Ljungdahl pathway (Marshall, Karst, et al., 2017).
We observed major differences between marine and lacustrine, and organic-rich and
organic-poor sediments in the types of possible carbon catabolisms, including carbohydrate
degradation, extracellular protein degradation, and fermentation pathways. Samples examined
here were deposited during multiple stages of the Baltic Sea development including: 1) the Baltic
Ice Lake stage, that occurred approximately 12,600 to 11,600 years BP and during which
sediments were predominantly glacial clays with relatively little TOC content (samples M59E-
61
67m, M63E-47m, and M65C-30m); 2) the Ancylus Lake stage, which was a freshwater stage
that occurred 10,700 to 10,200 years BP and had less TOC deposition and lower salinities than
the modern Baltic Sea (sample M65C-10m); and the 3) Littorina Sea stage, which was warm and
eutrophic leading to high sedimentation rates (samples M59E-15m, M65C-3m, and M63E-11m)
that developed by 8,500 years BP, as well as ‘modern’ samples (M59E-0.25m and M65C-0.25m)
(T. Andrén, Björck, E. Andrén, Conley, Zillén, and Anjar, 2011b).
The TC and TOC deposition has fluctuated significantly throughout the depositional
history of the Baltic Sea, and sediment OM contents range from 0.03% to 8.34%. As TOC is
vital to microbial community metabolism, we expected greater gene abundance and transcription
for organic degradation related genes in strata of high TOC content.
Relative percentages of carbohydrate degradation genes generally were higher in the high
organic non-glacial sediments compared to the glacial samples. However, gene abundances, most
notably within the CAZyme and peptidase genes, at site M60B often did not follow the
correlation between TOC content and gene abundances. This site is located outside of the Baltic
Sea Basin in the Kattegat region, and the samples examined here from 24 and 37 mbsf were
deposited during the Kattegat deglaciation c. 15,900 – 16,500 years BP (Friberg). During this
period, the Kattegat water column was marine-brackish due to inputs from the North Sea and
glacial meltwaters (Houmark-Nielsen and Henrik Kjær, 2003), whereas the Baltic Sea Basin was
cut off from marine inputs. Sediments collected at site M60B in the current study contained little
organic matter (~ 0.5 %), no methane accumulation, and still had millimolar concentrations of
pore water sulfate, indicating less microbial respiration occurring at this site compared to other
marine-influenced sediments in this study. Metagenomes from these samples contained larger
62
relative percentages of carbohydrate and protein-degrading genes than glacial samples from the
basin, even though the TOC contents were similar.
In statistical analyses, salinity of the samples was correlated with abundance of glycoside
hydrolases and peptidases. Studies have shown that the microbial community source dictates the
composition of sediment microbial communities (Hamdan et al., 2012; Walsh et al., 2016;
Starnawski et al., 2017). Melt water from continental glaciers can carry viable microorganisms
(Foght et al., 2004) and labile and recalcitrant organic matter to marine environments (Hood et
al., 2009). DNA can be preserved over long time scales in sediments (Torti et al., 2015;
Kirkpatrick et al., 2016), and sediment microbes can reflect previous depositional conditions
(Orsi et al., 2017). This could explain why samples with similar TOC contents but distinct
depositional histories (glacial lacustrine versus marine) carry different genetic signatures.
Potentially, the genetic potential found in the 60B samples were more reflective of past
depositional conditions, such as terrestrial glacial meltwater influxes or marine origins, as
opposed to present in situ activity.
These results further support emerging evidence that deep biosphere communities are the
result of seeding from the surface, and subsequent subseafloor selection (17, 91). In the Baltic
Sea, sediment organic matter content, pore water origin (reflected in salinities), and glacial
conditions were controlling factors in the microbial carbon mineralization potential. This adds to
a growing body of studies indicating that climate conditions impact sediment communities and
their potential biogeochemical cycling long past burial.
In summary, we examined microbial community potential and activity through DNA and
RNA sequencing to determine the types and relative abundances of carbon mineralization
pathways in Baltic Sea Basin sediments. We determined that in the organic-rich sediments, there
63
was potential for multiple metabolic strategies, including protein degradation, complex
carbohydrate usage, and methanogenesis. Furthermore, based on metatranscriptomic analyses,
these pathways were active. Fermentation, including alcohol production, was ubiquitous in all
sediments examined, and it appears that the ethanol production potential has been
underappreciated in sediments. Finally, while many functional gene abundances correlated with
carbon content and salinity of the sediment, the abundance of carbohydrate active enzymes in
organic-lean sediments indicates that in situ conditions are not the sole control on community
composition and that depositional conditions are important as well.
Acknowledgments
We thank the entire science party and crew of Integrated Ocean Drilling Program
Expedition 347: Baltic Sea Paleoenvironment and the Greatship Manisha, of the
Geomicrobiology of the Skagerrak and Kattegat cruise (2014), and AUBO16 cruise and R/V
Aurora. We greatly appreciate the help of Felix Beulig, André Pellerin, Alex Michaud, Gilad
Antler, Susanne Nielsen, and Aurora captain Torben Vang during sampling. We thank Megan
Mullis, Rachel Weisend, Morgan Sobol, and Pratixa Savalia for their assistance with laboratory
work, Mike Lee, Elaina Graham, and Benjamin Tully for their bioinformatics insight, and Annie
Rowe and Doug LaRowe for helpful discussions. This research was supported by National
Science Foundation Geology and Geophysics grant R01-1015237 to K.G.L., the Center for Dark
Energy Biosphere Investigations grant NSF OCE-1431598 to J.P.A., the Sloan Foundation Deep
Carbon Observatory’s Census of Deep Life, and Danish Center for Marine Research Grant
“Cryptic Biogeochemistry in the Bornholm Basin” to H.R. L.A.Z. was supported by a USSSP
Schlanger Fellowship. This is a C-DEBI contribution and also a Deep Carbon Observatory
contribution.
64

Figure 1. (a) Map of the samples locations in the Kattegat and Baltic Sea Basin (map adapted
from Andrén et al, 2015). (b) Chlorinity-based porewater salinity and (b) Total Organic Carbon
(TOC) content in percent dry weight and of sediments from IODP Expedition 347 (Andren et al,
2015). Sample depths are indicated in the space between panels b and c, and correspond to the
key in panel c.

65

Figure 2. Relative abundance of extracellular peptidases (listed by MEROPS or PFAM
nomenclature) in metagenomes and metatranscriptomes based on percent of reads mapped
relative to total reads, and normalized to reads per million (nrpm). Along the top of the heatmap
is the time period in which the samples were deposited (Holocene versus Late glacial)), and the
state of the Baltic Sea or Kattegat.

66

Figure 3. Extracellular carbohydrate degrading enzymes in the metagenomes and
metatranscriptomes, including Glycoside Hydrolases (GHs), Polysaccharide Lyases (PL), and
Carbohydrate Binding Modules (CBMs). Relative abundance is based on percent of reads
mapped to genes divided by the total number of reads in the sample, normalized to reads per
million (nprm). The type of CAZymes and putative target substrate(s) are listed along the y-axis,
the x-axis is arranged by sample type (RNA or DNA) and by the depositional times and
environments.
67

68
Figure 4. (a) Relative abundance of fermentation genes shown by sample in the metagenomes
(left) and fermentation gene transcripts in the metatranscriptomes (right) in normalized reads per
million (nrpm). (b) The relative abundance of pyruvate ferredoxin oxidoreductase and pyruvate
formate lyase genes and (c) of alcohol dehydrogenases in each sample.

69

Figure 5. Methane and other C1 metabolism genes assigned in mapped and unmapped reads.
Mapped reads were aligned to assembled genes assigned within putatively methanogenic
lineages.
70

71
Figure S1. Linear regressions or boxplots of abundant extracellular peptidases compared to
select environmental parameters. Blue dots represent samples deposited during the late
Pleistocene and are most glacially influenced, green samples were deposited during the Holocene
(i.e., the Ancylus Lake, Littorina Sea, and Baltic Sea phases of the Baltic’s history).

Figure S2. Relative abundance of carbohydrate degrading enzymes, which are involved in
smaller molecule degradation. Target substrates are labeled based on likely source of parent
substrates.

72

Figure S3. Linear regressions of abundant extracellular CAZyme groups compared to select
environmental parameters. Blue dots represent samples deposited during the late Pleistocene and
73
are most glacially influenced, green samples were deposited during the Holocene (i.e., the
Ancylus Lake, Littorina Sea, and Baltic Sea phases of the Baltic’s history).
74

75
Figure S4. Linear regressions of fermentation genes compared to select environmental
parameters. Blue dots represent samples deposited during the late Pleistocene and are most
glacially influenced, green samples were deposited during the Holocene (i.e. the Ancylus Lake,
Littorina Sea, and Baltic Sea phases of the Baltic’s history).

Figure S5. Canonical Correspondence Analyses of gene relative abundances between samples
based on functional classes: a) carbohydrate active enzymes; b) peptidases; and c) fermentation
genes.
76

Figure S6. Simplified stratographic columns of cores sampled during IODP X347. HM =
Holocene Marine; AL = Ancylus Lake; YS = Yoldia Sea; BIL = Baltic Ice Lake.

0
M59 M63 M60 M65
HM
AL
BIL
10
20
30
40
90
80
70
60
50
Clay Silt Sand Clay Silt Sand Clay Silt Sand Clay Silt Sand
HM
HM
AL
YS
BIL
Proglacial sand
BIL
AL
HM
Ice
influenced
deltaic
Marginal marine
meters below seafloor
Baltic Proper Kattegat
M59E-0.25m
M65C-0.25m
M59E-15m
M59E-67m
M63E-12m
M63E-47m
M65C-3m
M65C-10m
M65C-30m
M60B-24m
M60B-37m
77
Table 1. Characteristics of sediment samples here
Site Location
Water
Depth
Sample
Depth
Below
Seafloor
Depositional
conditions
Salinity
Total
carbon
Total
organic
carbon Alkalinity Methane Sulfate Formate Acetate
Lat, Long m m
Cl
-

based wt % wt % meq L
-1
mM mM µM µM
M59E
Little Belt
55°0.285ʹN,
10°6.499ʹE
37.1
0.25
Holocene
marine 23.30 6.40 5.40 - - 0.25 1.61 3.95
15
Holocene
marine 24.23 5.95 4.97 173.30 1.13 0.07 10.38 20.54
67
Glacial
lacustrine 7.48 2.17 0.91 4.40 1.76 0.01 0.74 24.04
M60B
Kattegat
56°37.204ʹN,
11°40.229ʹE
31.2
24
Marginal
marine 31.72 2.08 0.48 14.91 0.00 6.59 2.73 11.85
37
Marginal
marine 30.67 3.21 0.55 8.91 0.00 14.45 2.73 11.85
M63E
Landsort
Deep
58°37.330ʹN,
18°15.240ʹE
437.1
11
Holocene
marine 12.03 1.76 1.53 51.30 2.33 0.01 0.00 37.38
47
Glacial
lacustrine 1.67 0.70 0.55 17.23 9.40 0.02 0.38 22.73
M65C
Børnholm
Basin
55°28.084ʹΝ,
15°28.624ʹE
84.3
0.25
Holocene
marine 15 5* 4.99 - 0.19 2.3 2.92 1.80
3
Holocene
marine 15.52 4.16 3.67 38.00 10.10 0.03 2.55 34.90
10
Holocene
lacustrine/
marine
transition 12.84 0.88 0.97 32.70 9.30 0.07 2.32 19.50
30
Glacial
lacustrine 2.78 2.48 0.48 5.30 0.80 0.22 0.00 22.80
References: Andrén et al., 2015, Sci. Drilling; van Helmond et al., 2017; Hardisty et al., 2016. AJS; Jensen et al.,
2017. Boreas; Beulig et al. 2018. PNAS.

78
Supplemental Table 1. Sequencing and read statistics
Data Source
Sampling
campaign Site Sample
Depth
(m) Raw Post-QC
Post-
dereplication
Percent
mapping to
assembly
DNA
This paper
Skaggerak-
Kattegat Seabed
M59
SMT 0.25 49,239,081 46,783,878 46,058,076 26.41%
This paper IODP x347 5H2 15 29,293,749 28,328,313 26,195,655 35.95%
Marshall et
al., 2017 IODP x347 21H2 67 70,856,845 30,948,382 1,072,012 27.77%
Marshall et
al., 2017 IODP x347
M60
9H2 24 40,486,926 32,499,543 2,018,541 32.63%
Marshall et
al., 2017 IODP x347 13H2 37 54,835,820 41,888,865 4,363,223 32.98%
Marshall et
al., 2017 IODP x347
M63
6H2 12 38,620,470 25,846,898 4,309,404 24.27%
Marshall et
al., 2017 IODP x347 24H2 47 49,069,002 37,775,844 2,211,025 22.53%
This paper Bornholm Basin
M65
SMT 0.25 42,069,302 39,698,484 39,156,112 24.96%
This paper IODP x347 2H2 3 49,584,551 45,176,765 10,852,368 28.39%
This paper IODP x347 4H2 10 138,011,697 127,157,275 17,117,038 15.43%
Marshall et
al., 2017 IODP x347 10H2 30 51,710,778 23,750,244 2,115,020 24.64%
RNA
Zinke et al.,
2017 IODP x347
M59
5H2 15 16,968,921 15,232,832 10,908,171 28.36%
Zinke et al.,
2017 IODP x347
M63
6H2 12 17,904,120 16,857,431 12,645,054 24.02%

Assembly
Statistics No. of contigs
No. of bp in
contigs
Min bp
contigs
Max bp
contigs
Avg bp
contigs
N50 bp
contigs
Co-assembly 557,851 1,072,605,570 1,000 159,114 1,923 1,921

79
“What use are the best of arguments when they can be destroyed by force?”
-Jules Verne, 20,000 Leagues Under the Sea

80
4. Sediment microbial communities influenced by cool hydrothermal fluid
migration
In collaboration with Brandi Kiel Reese, James McManus, C. Geoffrey Wheat, Beth N. Orcutt,
and Jan P. Amend.
Abstract
Cool hydrothermal systems (CHS) are prevalent across the seafloor and discharge fluid
volumes that rival oceanic input from rivers, yet the microbial ecology of these systems are
poorly constrained. The Dorado Outcrop on the ridge flank of the Cocos Plate in the northeastern
tropical Pacific Ocean is the first confirmed CHS, discharging minimally altered < 15ºC fluid
from the shallow lithosphere through diffuse venting and advection. In this paper we characterize
the resident sediment microbial communities influenced by cool hydrothermal advection, which
is evident by elevated nitrate and oxygen concentrations. 16S rRNA gene sequencing revealed
that Thaumarchaea, Proteobacteria, and Planctomycetes were the most abundant phyla in all
sediments across the system regardless of influence from seepage. Members of the
Thaumarchaeota (Marine Group I), Alphaproteobacteria (Rhodospirallales), Nitrospirae,
Nitrospina, Acidobacteria, and Gemmatimonadetes were enriched in the sediments influenced by
CHS advection. Of the various geochemical parameters investigated, nitrate concentrations
correlated best with microbial community structure, indicating structuring based on seepage of
nitrate-rich fluids. A comparison of microbial communities from hydrothermal sediments,
seafloor basalts, and local seawater at Dorado showed differences that highlight the distinct niche
space in CHS. We conclude that cool hydrothermal venting at seafloor outcrops can alter the
local sedimentary oxidation-reduction pathways, which in turn influences the microbial
communities within the fluid discharge affected sediment, and that communities from Dorado
81
differ from those at previously characterized, warmer CHS sediments. However, these
communities are similar to previously characterized deep sea sediments, such as sediments from
the South Pacific Gyre.
Introduction
Ridge flank hydrothermal systems are globally widespread and responsible for over two
thirds of marine hydrothermal heat flux (C. A. Stein and S. Stein, 1994). A significant amount of
this flux is proposed to be through low temperature fluids at Cool Hydrothermal Systems
(CHSs). These systems bring cold seawater into the shallow lithosphere, circulate this fluid over
short time scales of years to tens of years, and vent cool (< 20°C) fluid into the ocean through
seafloor outcrops (Fisher et al., 2003; Wheat and Fisher, 2008). More than 25 million seamounts
and even more smaller basaltic outcrops that can facilitate CHS fluid flow, are predicted to exist
in the ocean (Wheat et al., 2003)., Importantly, CHS systems discharge fluid volumes that rival
oceanic input from rivers, with an estimated 10
14
kg yr
-1
of fluid flowing through just the ~15,000
largest seamounts (Harris et al., 2004). The fluid movement facilitated by CHS are responsible
for oceanic removal of phosphate at approximately of 15% of the global riverine flux (Wheat et
al., 2017).
Advection and diffusion drive cool oxic fluids upwards into the thin sediment covering
parts of outcrops, which serve as discharge sites for CHSs. This fluid advection can elevate
porefluid nitrate and oxygen concentrations relative to sediment without fluid seepage (Wheat
and Fisher, 2008; Wheat et al., 2017) and impact mineral composition (Bodeï et al., 2008). The
delivery of oxidants from the crustal fluid makes high-energy electron acceptors, such as oxygen
and nitrate, available for microbial metabolism at sediment depths where they would otherwise
be depleted. Similar phenomenon was recently observed in sediment overlying the flank of the
82
Mid-Atlantic Ridge, where diffusion of oxygen and nitrate into basal sediment ponded between
crustal exposure stimulates a nitrogen-cycling microbial community (Reese et al. 2018).
Although the global significance of CHS to heat and chemical exchange has been
demonstrated, there is a lack of understanding of how these fluid flows impact the structure of
sediment microbial communities or vice versa. Initial studies have demonstrated that oxic fluid
flux changes the microbial community, but currently these studies have been performed at a
limited numbers of sites, including much thicker sediment columns than at Dorado (e.g. (Reese,
Zinke, Sobol, Doug E LaRowe, et al., 2018) or at much greater temperature of source fluid (e.g.
(Huber et al., 2006). Hence, further characterization of CHS sites with shallow sediment cover
(e.g., direct access of fluids to the overlying water) is necessary to determine how CHS fluids are
impacting the associated sediment chemistry and thereby the microbial communities.
Here, we present a characterization of sediment microbial communities from the Dorado
Outcrop (Figures 1, 2), located at approximately 3,000 meters water depth on a 20 to 23-million-
year-old region of the Cocos Plate (Wheat and Fisher, 2008). This outcrop is the first confirmed
site of cool (<15
o
C) hydrothermal flow from a CHS (Wheat and Fisher 2008, Hutnak et al.
2008), and recent investigations confirmed that venting fluids are replete with oxygen (<55 m M)
and nitrate (<38m M; Wheat et al., 2017). Sediment porewater profiles confirm the upward flux
of oxygen and nitrate into basal sediment layers surrounding the outcrop, a signature of CHS
fluid influence in the sediment (Wheat and Fisher, 2008; Wheat et al., 2017). Likewise, previous
investigation of Dorado Outcrop sediment profiles of dissolved and solid phase manganese
indicates oxidizing conditions in sediment influenced by CHS (i.e., less dissolved Mn, greater
solid phase Mn), whereas background sediment not affected by CHS flow had more reducing
conditions (i.e., greater dissolved Mn, less solid phase Mn; Wheat and Fisher 2008, Bodeï et al.,
83
2008). In the background samples, solid manganese (i.e. manganese oxides) reduction by
microorganisms resulted in the accumulation of manganese in porewaters and depletion of solid
phase manganese. The hydrothermal sediments did not show this pattern, indicating a more
oxidizing sediment column and reflecting manganese concentrations similar to crustal fluid as
opposed to a typical diagenetic pattern (Wheat and Fisher, 2008). A recent study of basalts on the
outcrop did not indicate a clear signature of CHS flow on the structure of microbial communities
on the exterior of the basalts, however (Lee et al., 2016). These prior characterizations make the
Dorado Outcrop an ideal location to examine the potential influence of CHS seepage on
sediment microbial communities.
The objectives of this study were to determine (1) the composition of microbial
communities present in CHS sediments, (2) how hydrothermally affected sediment communities
differed from those in nearby background sediments, seafloors basalts, and bottom seawater; (3)
if geochemical changes associated with CHSs impact putative microbial metabolic potential at
these sites; and (4) to determine how Dorado compares to other deep sea sites. We hypothesized
that the elevated concentration of oxidized compounds (e.g., oxygen, nitrate) in hydrothermal
sediments would alter the overall community composition.
Methods
Site description
The Dorado Outcrop is located in the Eastern Tropical Pacific Ocean (9°5’N, 87°5’W,
Fig 1A) on a swath of seafloor derived from the East Pacific Rise (Fisher et al., 2003). The
outcrop rises approximately 100 – 150 m above the surrounding seafloor, is 2 km long, and 0.5
km wide (Figure 2D; Wheat et al., 2017). The sediment package surrounding the outcrop is
approximately 200 – 400 m thick (Fig 1).
84
Sample collection
Sediment samples were collected from the Dorado Outcrop December 4-10, 2014 during
cruise AT26-24 aboard the R/V Atlantis. Sediment was retrieved via push core (10-28 cm length)
using the DSV Alvin during dive numbers 4777, 4780, 4782, and 4783 (Fig. 2D). Cores for
geochemistry and microbiology were taken adjacent to each other, with all related cores taken
within half a meter from each other. These cores will be referred to hereafter as Push Core (PC)
1-9 (Figure 2, Table 1). Based on location and geochemical characterization described below,
samples from PC1, 2, 5, 6, and 8 are collectively referred to as ‘hydrothermal’; samples from
PC4, 7, and 9 as ‘intermediate’; and samples from PC3 as ‘background’.
Once shipboard, cores were examined for cracks or seawater intrusion. Cores with no
visible evidence of seawater intrusion were stored vertically at 4°C until porefluid extraction or
sediment sectioning could begin, usually within 12 hours of sampling. Dissolved oxygen data
were obtained by microsensor measurements through side ports of companion cores (Wheat et
al., 2017). Porefluids to measure dissolved nitrate and manganese were collected with Rhizons
(Rhizosphere Research Products), which were inserted through pre-drilled side ports of
individual sediment cores (e.g., Ziebis et al., 2012). Samples for bulk manganese compositions
were determined from 1-cm to 10-cm intervals that were selected at sea and refrigerated for
further shore-based handling and analysis. Sediment samples for microbial analysis were taken
from the center of the cores using sterile plastic syringes that had been cut off at one end,
wrapped in foil, and autoclaved before use. Samples were collected in 6 cm intervals from each
sediment core and immediately frozen at –70°C. Samples were shipped on dry ice to the
University of Southern California and stored at –80°C until DNA extraction.

85
Nitrate and manganese measurement

Dissolved nitrate and manganese were determined using colorimetric or ICP emission
techniques, respectively, that were identical to those used to analyze discrete samples of
discharging fluids (Wheat et al., 2017). Dithionite extractable manganese was determined ashore
on freeze-dried sediment that was crushed with an agate mortar and pestle. The dithionite
extractable manganese concentration was determined using a single step chemical leach that was
designed to remove labile metals from sediment (Mehra and Jackson, 1960, Roy et al., 2013;
Murray et al., 2016). Briefly, 0.25 g of sediment was heated at 60°C for 4 hours in a buffered
sodium dithionite solution with agitation via vortex every 15 minutes. Samples were then cooled
before centrifuging at 4000 RPM for 5 minutes. The supernatant was decanted, diluted (1:20
with 18 megaohm water) and analyzed for manganese with a Perkin Elmer Atomic Absorption
Spectrometer AAnalyst 700, or on an Agilient Technologies 700 series inductively coupled
plasma optical emission spectrometer (ICP-OES). Dissolved and solid phase geochemical data
are available from the Biological and Chemical Oceanography Data Management Office (BCO-
DMO) database under project number 627844.
DNA extraction
DNA was extracted in triplicate from stored, frozen sediment cores (3-4 and 9-10 cm
below seafloor, cmbsf). For each replicate, approximately one gram of sediment was divided
between two screw-cap 2-mL tubes. DNA was extracted in a UV-sterilized clean hood using the
FastDNA SPIN Kit (MP Biomedicals, Santa Ana, CA) using two reactions for each replicate.
The two reactions for each replicate were combined during the SPIN filter step. Biological
replicates were not combined. DNA was eluted in 50 µL of molecular biology grade sterile
water. Blank extraction controls with no sample added were run alongside each extraction to
86
verify sterility. Resulting extractions were quantified using the Qubit HS dsDNA Assay on a
Qubit 2.0 Fluorometer (Life Technologies, Carlsbad, CA) following manufacturer protocols.
DNA sequencing
A total of 54 samples (triplicates of two depths from nine push cores) and three extraction
blanks were sequenced at the Molecular Research DNA Lab (Shallowater, TX, USA). The V4
region of the 16S rRNA gene was amplified using the Earth Microbiome Project universal 515F
(5ʹ-GTG CCA GCM GCC GCG GTA A-3’) and 806R (5ʹ-GGA CTA CHV GGG TWT CTA
AT-3’) primers (Caporaso et al., 2012). The forward primers included 8 nucleotide barcodes.
Libraries were created through PCR with HotStarTap Plus (Qiagen, Germantown, MD) using the
following protocol: 94°C for 3 minutes; 28 cycles of 94°C for 30 seconds, 53°C for 40 seconds,
and 72°C for 1 minute; and 72°C for 5 minutes. Amplified DNA was pooled in approximately
equimolar concentrations and purified using Ampure XP beads (Beckman-Coulter, Indianapolis,
IN). Due to low initial DNA concentration, extraction blank libraries did not produce signal after
the initial 28-cycle amplification, so the PCR was extended to 40 cycles prior to sequencing.
DNA was sequenced on an Illumina MiSeq platform using Illumina TruSeq (Illumina, Inc., San
Diego, CA) chemistry with 2 x 250 base pair chemistry. Resulting sequence data were trimmed
of barcodes and low-quality sequences using a quality cutoff of 25, and sequence read pairs were
merged by the sequencing facility, resulting in an average sequence length of 299 base pairs.
Sequence Analysis
Sequence primers were removed using Cutadapt (Marcel Martin, 2011). Sequence files
containing all sequences were split into individual files using the split_libraries_fastq.py and
split_sequence_file_on_sample_ids.py commands in QIIME (Caporaso, Kuczynski, Stombaugh,
Bittinger, Bushman, Costello, Fierer, Pe a, et al., 2010), including a minimum quality score
87
threshold of 25 for all sequences. Merged sequences were processed in Divisive Amplicon
Denoising Algorithm 2 (DADA2) v1.6 following the described protocols (Benjamin J Callahan
et al., 2016; Ben J Callahan et al., 2016) implemented in R version 3.4.1 (R Core Team, 2016),
and the following analyses were performed in DADA2 unless otherwise stated. Sequences
primers were removed using Cutadapt (Marcel Martin, 2011). Sequence files containing all
sequences were split into individual files using the split_libraries_fastq.py and
split_sequence_file_on_sample_ids.py commands in QIIME (Caporaso, Kuczynski, Stombaugh,
Bittinger, Bushman, Costello, Fierer, Pe a, et al., 2010), including a minimum quality score
threshold of 25 for all sequences. These sequences were imported into DADA2, where they were
further filtered and trimmed to a length of 240 base pairs following the suggested DADA2
workflow. Amplicon Sequence Variants (AVSs), which are analogous to 100% sequence
similarity, were inferred from sequence data (Ben J Callahan et al., 2016). Sequences
representative of each ASV were chimera checked and chimeric ASV were removed.
Sourcetracker was used to identify potential ASVs sourced from lab or DNA extraction kit
contamination, and these OTUs were removed from samples before further downstream analyses
(Knights et al., 2011). Taxonomy (phylum through genus levels) was assigned to ASVs using the
Ribosomal Database Project’s naïve Bayesian classifier (Cole et al., 2009) with the Silva v128
database as the reference (Pruesse et al., 2007). Sequences from the current study were deposited
into the National Center for Biotechnology Information (NCBI) Sequence Read Archive (SRA)
database under project PRJNA433243.
Dorado Outcrop bottom seawater and seafloor basalt sequences collected in 2013 were
included in sequence processing in order to assess potential contamination of sediment by
seawater intrusion and similarities between CHS impacted sample types (i.e., basalt and
88
sediment). Collection, extraction, and sequencing details are included in Lee et al. (2015). These
samples were extracted using the same brand and type of extraction kit (MP Biomedicals
FastDNA Spin Kit) and were sequenced with the same primer set (EMP 515F - 806R) at
Molecular Research DNA Lab (Lee et al., 2015). These sequences are accessible at NCBI SRA
project SRP063681.
Sequences from the Clarion Clipperton Zone were downloaded from NCBI SRA project
SRP057408. These sequences were collected and sequenced as described previously (Shulse et
al., 2017). DNA extraction and sequencing protocols were similar to those described here: DNA
extraction was performed with the FastDNA Spin Kit, and sequencing was performed with the
EMP 515F – 806R primer set on the MiSeq platform.
Statistics
Statistical analyses between the 16S rRNA gene sequences (abundance and taxonomic
assignment) and sediment geochemical data were performed using the phyloseq v1.22.3
(McMurdie and Holmes, 2013), DESeq2 v3.6 (Love et al., 2014), and Vegan v2.4.4 packages in
R (Dixon, 2003). Plots were generated using ggplot2 (Wickham, 2011) and base R. Taxonomic
assignment bar charts were generated in phyloseq using ggplots2. Relative abundances of data
were determined by dividing the number of sequences for the ASV or taxon by the total number
of quality controlled sequences for each sample. Ordinations (including NMDS, PCoA, and
CCA) were generated using non-rarefied log transformed or rank abundance data (Ben J
Callahan et al., 2016) and the Bray-Curtis dissimilarity index (Bray and Curtis, 1957).
Ordinations were plotted using ggplot2 v2.2.1 (Wickham, 2011). Differential abundance data
and statistics were generated on ASVs using the DESeq2 package using the Wald test statistic
and parametric fit (Love et al., 2014). Correlations between tested parameters and OTUs were
89
considered significant when the p-value was less than 0.05. Additional statistical testing,
including student’s t-tests, linear regressions, and NP-MANOVA analyses were conducted in R
using the vegan package (Dixon, 2003). Violin plots were made in R by agglomerating OTUs at
the order level and plotting them based on abundance. PCoA analysis with weighted Unifrac was
produced in QIIME (Caporaso, Kuczynski, Stombaugh, Bittinger, Bushman, Costello, Fierer,
Pena, et al., 2010). Alpha diversity was calculated in phyloseq using rarefied data (Table S5).
Results
Sediment geochemistry
Dissolved nitrate, oxygen, and manganese depth profiles differed among core
classifications (Fig. 1B, C, and E). The background sediment core (PC3) had very low nitrate
concentrations (Fig.1B). The hydrothermal cores had nitrate concentrations (~30-40 µmol kg
-1
)
similar to those measured in bottom seawater, a pattern previously interpreted as an indication of
CHS advection (Wheat and Fisher, 2008). Nitrate concentrations in the three intermediate cores
(6.51 – 44.31 µmol kg
-1
) were between those in the hydrothermal and background cores. The
background core (PC3) was depleted in dissolved oxygen and ranged between 0 and 1.1 µmol
kg
-1
(Fig. 1C). Two hydrothermal cores had measureable oxygen throughout most of the
sediment column (PC8 and PC6). Some profiles also show an increase in oxygen concentration
with increasing sediment depth near the sediment-basement interface (Fig. 1C). The porefluid
manganese concentration was greater in the background sediment core (PC3) than in the
hydrothermal and intermediate cores, where concentrations were typically below detection limit
(Fig. 1E). This pattern is consistent with prior observations where dissolved Mn is lower in
sediment influenced by CHS seepage (Wheat and Fisher, 2008). Solid phase manganese
concentrations in the cores varied between below detection limit (typically less than or equal to
90
0.03 wt %) and 1.7 wt % of dry sediment (Fig. 1F). The background core showed solid phase
manganese depletion in the first 10 cmbsf. Hydrothermal and intermediate core profiles did not
show a clear pattern, but generally solid manganese decreased with depth.
16S rRNA gene sequencing, clustering, and taxonomic assignment
Quality filtering of 4,659,316 paired-end sequences resulted in 4,459,234 sequences. The
number of sequences per sample ranged from 6,786 to 127,268, with a mean of 78,232 ± 24,884
sequences (Supplemental Table S1). Clustering and chimera checking in DADA2 (Benjamin J
Callahan et al., 2016) produced 19,577 Amplicon Sequence Variants (ASVs) at the 100%
sequence similarity level. Sourcetracker analysis showed there was little (<0.01%) taxonomic
overlap between the ASVs from the sediment microbial communities. Sediments, blanks, and
bottom water samples did not group together (bottom water data from Lee et al. 2015) (Fig. S2).
Clustering of triplicate sample sequence libraries using the Bray-Curtis dissimilarity index
revealed consistency (clustering) of microbial community structure amongst replicates for most
samples, with only two samples not clustering together (Fig. S3), indicating repeatable, quality
sequencing per sample.
The major phyla in all sediment samples on and near Dorado Outcrop were
Proteobacteria (29.9 ± 6.1%), Planctomycetes (19.9 ± 4.5%), Thaumarchaeota (12.6 ± 7.7%),
and Chloroflexi (7.8 ± 3.1%); an average of 6 ± 3% of sequences could not be assigned to a
taxonomic group at the phylum level (Fig. 3). Within the Proteobacteria, the Delta class was the
most abundant (12.8 ± 3% of total community), followed by the Alpha (8.5 ± 3.2%) and Gamma
(7.7 ± 3.4%) classes (Figs 3, S4-6) . The Alphaproteobacteria were mostly within unclassified
genera in the Rhodospirillaceae family (77.3% of Alphaproteobacteria), and the
Gammaproteobacteria were mostly assigned as genera in the Xanthomonadales order (67.9% of
91
Gammaproteobacteria). The most abundant classified genera were the Candidatus Scalindua (9.1
± 4.3%), the Candidatus Nitrosopumilales (5.0 ± 3.8%), the Urania-1B-19 marine sediment
group genus within the Physcisphaereae family (2.4 ± 0.9%), Nitrospira (1.4 ± 1.5%), the H16
genus within the Desulfurellaceae family (Deltaproteobacteria) (1.3 ± 0.8%), the Pir4 lineage
within the Planctomycetaceae (1.1 ± 0.5%), Nitrospina (0.8 ± 0.9%), Candidatus Omnitrophus
(0.3 ± 0.2%), and Nitrosomonas (0.3 ± 0.2%) (Fig. S5).
Some of the most abundant ASVs showed enrichment in the hydrothermal samples
relative to background samples (Fig 4). These included ASVs assigned within the Nitrospina
(ASVs 82 and 92; average abundance of 0.3 – 0.4% of total sequences in HF, 0.00% in Int, and <
0.01% in BG), Nitrospira (ASVs 12 and 23; 0.3 – 0.5% HF, £ 0.02% Int, and 0.02 – 0.03% BG),
Gemmatimonadetes class BD2-11 (ASV93; 0.2% HF, 0.01% Int, and 0.02% BG),
Alphaproteobacteria family Rhodospirillaceae (ASV 3; 1.5% in HF, 0.6% in Int, and 0.1% in
BG), and SAR202 Chloroflexi (ASVs 235 and 402; 0.1 – 1.5% HF, < 0.1% Int, and 0% BG).
Thaumarchaea assigned ASVs were notably more abundant in the HF compared to the BG
samples (Fig. 4E).
ASVs which were generally more abundant in the BG and/or Int samples were assigned
as Candidatus Scalindua (including ASVs 1 and 2; 0.5 – 1.5% HF, 3.6 – 4.6% Int, and 3.2 –
4.2% BG), Ignavibacteria family PHOS-HE36 (ASV29; 0.2% HF, 0.5% Int, and 0.7% BG),
Deltaproteobacteria family Syntrophobacteraceae (ASV 38, 238, and 260; £ 0.1% HF, 0.1 –
0.4% Int, and 0.1 – 0.7% BG), Aminicenantes ASVs (ASV87 and 232; 0% HF, £ 0.3% Int, and
0.3 – 0.5% BG), Chloroflexi family Anaerolineaceae (ASVs 47; 0.0% HF, 0.3% Int, and 0.8%
BG), and the phylum Aminicenantes (ASV87; 0.0% HF, 0.3% Int, and 0.3% BG) (Fig. 5). Of the
most abundant ASVs, many ASVs (68 of the 100 most abundant) were either present in
92
relatively similar abundances across samples (e.g. Actinobacteria and Gammaproteobacteria
ASVs) or showed no apparent pattern (Fig. S7).
Differential abundance testing using DESeq2 was performed on community composition
data to determine which microbial taxa were significantly different among sample types (e.g.,
hydrothermal, intermediate, and background). Taxa whose relative abundances increased in
hydrothermal samples included Acidobacteria subgroup 21, Nitrospinae, Nitrospirae,
Rhodospirillales, and Gemmatimonadetes bacteria, and Marine Group I Thaumarchaeota (Figs.
3, S6A-F). Student’s t-tests showed no significant differences in relative abundance of these taxa
with depth in their classification (assuming a significance threshold of p < 0.05). Taxa whose
relative abundance decreased in hydrothermal samples included Deltaproteobacteria groups,
Chloroflexi, Ignavibacteriae, and some Planctomycetes classes (Figs. 4, S6G-L).
When samples were classified by nitrate concentrations, ASVs within 19 phyla were
found to differ significantly (p < 0.05) between samples (Table S2). Taxa that were positively
associated with nitrate concentration included Thaumarchaeaota, Nitrosomonadales (within
Betaproteobacteria), and Rhodospirillales (within Alphaproteobacteria). The ASVs that were
significantly negatively associated with nitrate included those assigned to Anaerolineae and
Dehalococcoides classes within the Chloroflexi phylum and Planctomycetes (Table S2).
Microbial community analysis
Ordination analyses were used to determine if sample types (i.e. hydrothermal,
intermediate, background) formed distinct groupings based on community composition.
Principal Coordinate Analysis (PCoA) showed that background samples grouped separately from
hydrothermal samples (Fig. 6A). PCoA also showed that within sample types, some delineation
of samples occurred. Sample groupings generally corresponded to nitrate concentrations when
93
these were overlain onto the PCoA (Fig. 6B). Canonical correspondence analysis (CCA)
incorporated community composition, nitrate concentrations, and solid phase manganese content
of samples into ordinations. As in the PCoA, hydrothermal communities were more similar to
each other than they were to background communities (Fig. 6C). Background communities from
9-10 cmbsf ordinated the furthest from other communities, and intermediate communities
ordinated in between the background and hydrothermal communities. Within the sample types,
samples grouped by depth, with the most pronounced trend in the hydrothermal samples (Fig.
6C). An exception was sample PC1 9-10 cmbsf, a hydrothermal sample that ordinated most
closely to intermediate 3-4 cmbsf samples.
Sediment samples were compared to seafloor basalts and bottom seawater described in
Lee et al. (2015). Sediment communities were significantly different from basalt and seawater
communities (p < 0.05), and coordination analyses showed grouping of the communities by
sample types (Fig. 6D). Notable differences among the communities were in the relative
abundances of Bacteroidetes (1% and 2% of sediment and basalt, respectively, versus 9% of
seawater), Chloroflexi (9% average of community between all sediments versus 2% average in
basalts and seawater), and Planctomycetes (average 22% of sediment communities and only 6%
and 4% of basalt and seawater communities, respectively) (Fig. 3).
Dorado Outcrop hydrothermal sediment communities were further compared to Clarion
Clipperton Zone (Pacific Ocean) sediment communities (Fig. 7). The most abundant classified
genera (> 0.1% of total community) were similar. Notably, Candidatus Nitrosopumilus,
Planctomycetes lineages (including the Urania-1B-19, Rhodopirellula, Planctomyces, and Pir4
genera), the Deltaproteobacteria H16 genus, Candidatus Omnitrophus, Nitrospira, and
Nitrospina were present in both sets of samples. Furthermore, the relative abundances of
94
Candidatus Nitrosopumilus, Deltaproteobacteria H16, and Candidatus Omnitrophus were not
significantly different (t-test, p > 0.05). However, some of the abundant genera were
significantly enriched (t-test, p < 0.05) in Dorado hydrothermal sediments. These genera
included Nitrospira, Nitrospina, and all of the Planctomycetes genera. Abundances of the
Candidatus Scalindua showed the largest difference in abundances between the sites, with
approximately 6.5% of Dorado hydrothermal sequences assigned to the genus. In contrast, 0.02%
of Clarion Clipperton Zone sequences were assigned to this genus.
Discussion
CHS-impacted sediment microbial communities
Whereas many hot hydrothermal sediment microbial communities have been described in
detail (Moyer et al., 1995; Teske et al., 2002; Cerqueira et al., 2015; Dowell et al., 2016),
sediment communities impacted by discharge of cool, oxygen- and nitrate-enriched basaltic
formation fluids are less well understood. To determine how these communities evolve in
response to fluid migration, we need to first understand microbial community structures in
‘typical background (oligotrophic) sediment, in bottom seawater, in the fluids that migrate
through ocean crust and discharge at outcrops, and on the basaltic rocks that the fluids are
exposed to.
In this study, we determined that background and CHS-influenced sediment at the
Dorado Outcrop (the first confirmed site of significant CHS discharge (Wheat et al., 2017)) are
dominated by the many of the same taxa (Fig. 3, Fig. S7), but within this overall similarity, some
taxa were correlated with CHS influence (Figs. 4, 5). Some of these taxonomic groups (the
genera Nitrospina, Nitrospira, and Candidatus Nitrosopumilus) are known to be involved in oxic
nitrogen cycling processes (Könneke et al., 2005; Lücker et al., 2010; Walker et al., 2010;
95
Lücker, Nowka, Rattei, Spieck, and Daims, 2013a; Rosenberg et al., 2014), suggesting that the
advection of nitrate and/or oxygen is a strong driver favoring the stimulation of microbial groups
that can perform such metabolic processes.
Dorado Outcrop sediments exhibit community compositions similar to those in other
deep sea sediments. For example, the “North Pond” site on the western flank of the Mid-Atlantic
Ridge has similar circulation of oxygen-enriched fluids within basement underneath the sediment
(Ziebis, McManus, Ferdelman, Schmidt-Schierhorn, Bach, Muratli, Edwards, and Villinger,
2012b; Orcutt, Wheat, et al., 2013). The sediment temperatures were similar at these two
locations, and both had notable abundances of Alpha-, Beta-, and Gammaproteobacteria (Reese,
Zinke, Sobol, Douglas E LaRowe, et al., 2018). These classes were abundant in the Dorado
sediment communities, though lower taxonomic levels were less similar. For instance, North
Pond contained abundant sequences from the genuera Brevundimonas (Alphaproteobacteria),
Achromobacter and Delftia (Betaproteobacteria), and Pseudomonas (Gammaproteobacteria).
Dorado Outcrop sediments did not contain any significant portions (more than 1% abundance) of
these taxa. Additionally, numerous diatom chloroplast sequences were preserved at North Pond,
which were not found here.
However, as found in Dorado sediment, North Pond sediment microbial communities
correlated with nitrate concentrations (Reese et al. 2018). In another example, microbial
communities in shallow oligotrophic sediment of the South Pacific Gyre (Durbin and Teske,
2011) display similarities to those in Dorado sediment. Namely, in South Pacific Gyre oxic
zones, Planctomycetes (including Candidatus Scalindua), Chloroflexi (SAR-202,
Dehalococcoides, Anaerolineae), Alphaproteobacteria (Rhodospiralleles), Gammaproteobacteria,
and Gemmatimonadetes are present, and the archaeal community was dominated by MG-1
96
Archaea (Durbin and Teske, 2011), which is similar to the oxic communities here. With
increasing depth in the South Pacific Gyre (SPG) sediment, oxygen and nitrate concentrations
decreased, and the relative abundances of Chloroflexi (specifically Dehalococcoides and
Anaerolinae) and Plantomycetes increased; this pattern is similar to that observed in Dorado
Outcrop sediment (Fig. 4, Fig S5). Another striking similarity between Dorado Outcrop and SPG
sediments was the disappearance of MG-1 Thaumarchaea in the oxygen and nitrate depleted
SPG sediments. We observed a similar trend in Dorado Outcrop sediment. MG-1 sequences were
still present in the anoxic BG sediments, but at a much lower abundance compared to the nitrate
rich hydrothermal sediments.
Lastly, dominant taxa in the top 10 cmbsf of sediments from the Clarion-Clipperton Zone
in the Eastern North Pacific were identified as Gammaproteobacteria, Alphaproteobacteria
(specifically Rhodospirillaceae), Deltaproteobacteria, Planctomycetes (Phycispharae), and
Thaumarchaea (including Candidatus Nitrosopumilis) (Shulse et al., 2017). These taxa were also
among the most abundant taxa in hydrothermal sediments at Dorado, and analyses revealed
similarities on the genus level (Figs. 7), and often these community members were of similar
relative abundances. Interestingly, Candidatus Scalindua was significantly enriched in Dorado
Outrcrop sediments, as were several other genera associated with nitrogen cycling (Nitrospira
and Nitrospina). However, Candidatus Scalindua has been observed in clone libraries throughout
marine sediments (Penton et al., 2006), including deep sea sediments in the South China Sea
(Hong et al., 2011). These results indicate that Dorado Outcrop hydrothermal sediments are
similar to other deep sea (non-hydrothermal) sediments, though with crucial differences in some
abundant community members.
97
By contrast, microbial communities in Dorado sediment are less similar to those in Juan
de Fuca Ridge flank sediment. Like Dorado sediments, outcrop sediments in the Juan de Fuca
system are relatively cool (7 – 64°C at the sediment basement interface) and are influenced by
migration of basement fluids. However, critical differences between these systems include
system age (~3.5 million years at Baby Bare versus 23 million years old at Dorado) and
hydrothermal fluid geochemistry – fluids collected at Baby Bare are anoxic and nitrate-depleted.
These differences are apparently reflected in the microbial community composition. Chloroflexi
and Deltaproteobacteria not closely related to clades found here, Epsilonproteobacteria,
and Aminicenantes (OP8) phyla were the most dominant taxa in Baby Bare sediment (Huber et
al., 2006). These clades were not as prominent in Dorado hydrothermal sediments, though
interestingly, one of the archaeal clones retrieved can be classified as Candidatus
Nitrosopumilus, so there are some similarities between the sites.
Sediments at Dorado contained microbial communities that were also distinct from
nearby (within a ~2 km
2
area) seafloor-exposed basalts and bottom seawater (Fig. 3, Fig. 6D). It
appears substrate plays an important role in selection of these communities, as sediments,
basalts, and seawater were all distinct at the Dorado Outcrop. Amongst sediments here, nitrate
and oxygen exert a significant selective force. In total, sediment chemistry on Dorado is driven
by CHS conditions, and CHS associated chemistry impacts community structure and succession.
Despite the differences between CHS sediments at Dorado and non-hydrothermal sites (notably
active flux of crustal fluid through sediments versus diffusion controlled systems), the
hydrothermal sediment communities on Dorado are similar to other non-hydrothermal deep sea
sediments in the Pacific.

98
Venting causes changes in putative microbial metabolic potential
Sediment microbial community succession is often driven by geochemical changes
associated with reduction of terminal electron acceptors (Nealson, 1997). Porefluids from the
background site are depleted in oxygen and nitrate, compared to bottom water and dissolved
manganese concentrations are enriched. In contrast to this background site, sediments that are
influenced by fluid advection processes near the expulsion of hydrothermal fluids generates
profiles of oxygen, nitrate, and manganese that deviate from those in background sediments
(Wheat and Fisher, 2008).
Nitrogen respiration has increasingly been recognized as an important metabolism in
oligotrophic deep sea sediments (Orcutt, LaRowe, et al., 2013; Wankel et al., 2015; Reese,
Zinke, Sobol, Douglas E LaRowe, et al., 2018). At Dorado Outcrop, taxonomic analyses of
resident microbial communities revealed lineages in CHS sediments related to known nitrogen
cycling organisms. However, the inference of function from taxonomy remains tenuous due to
their capability to perform multiple functions and processes like horizontal gene transfer, which
is extensive in prokaryotes (Koonin et al., 2001). We recommend these metabolic functions as
putative or hypothetical functions rather than absolute functions, and here we will focus on
genera with cultured representatives that have been previously identified to have metabolisms
related to nitrogen cycling, which have been found to be typically conserved phylogenetically
(Martiny et al., 2013).
Hydrothermal samples contained significantly larger percentages of Candidatus
Nitrosopumilus-related sequences than did intermediate and background, especially 9-10 cmbsf.
Cultured representatives of these Thaumarchaea are known to aerobically oxidize ammonium to
nitrite (Könneke et al., 2005; Walker et al., 2010), although recent studies have revealed some
99
Thaumarchaea may be heterotrophic or mixotrophic (Swan et al., 2014), utilize urea in
respiration (Alonso-Sáez et al., 2012), or anaerobically oxidize ammonium to nitrite (Jørgensen
et al., 2012). Sequences also were assigned as Nitrosospira, a genus of Betaproteobacteria known
to oxidize ammonium to nitrite (Prosser et al., 2014), including one OTU, which was among the
most abundant OTUs here (Figs 7, S5). OTUs classified within the Nitrospira and Nitrospina
genera were found in these sediments as well, pointing to aerobic oxidation of nitrite to nitrate
(Lücker et al., 2010; Lücker, Nowka, Rattei, Spieck, and Daims, 2013b). In addition to the
advection of nitrate rich hydrothermal fluid providing elevated nitrate, some nitrate might be
produced in situ through the action of putative ammonium and nitrite oxidizers. In contrast, the
relative abundance Candidatus Scalindua-related sequences in the intermediate and background
sediments could indicate anaerobic ammonium oxidation (anammox) is a critical nitrogen
cycling process in some sediments on and near Dorado. This notion is in accordance with
findings in other anoxic or suboxic sediments, which found that anammox can account for up to
80% of dinitrogen production (Devol, 2015).
This study is the first to characterize microbial communities in ridge flank sediments
associated with active cool hydrothermal discharge. Oxygen- and nitrate-enriched crustal fluid
advection established geochemical conditions, especially in deeper sediments, that favored
microbial communities different from those in background sediment. CHS-influenced sediment
communities were diverse, and contained Thaumarchaea, Proteobacteria, Planctomycetes, and
Chloroflexi related sequences. These communities were similar to those from other oxic cold
surficial marine sediments, and included large percentages of taxa related to known aerobic
nitrogen cycling organisms.

100

Acknowledgements
We would like to thank the entire Dorado Outcrop scientific party, the R/V Atlantis crew, and the
HOV Alvin crew on cruise AT26-24. We especially thank Annie Hartwell, Natalie Murray, Bo
Montagne, and Trevor Fournier for their assistance with fluid sampling, Mike Lee and Beate
Kraft for their assistance with sediment sampling, and Samuel Hulme and Chris Trabaol for
assistance with mapping. We thank Pratixa Savalia and Juan Orantes with laboratory assistance.
We also thank the comments of two anonymous reviewers. Preliminary sequencing for this work
was provided by the Deep Carbon Observatory (DCO) Census of Deep Life, which is funded by
the Alfred P. Sloan Foundation. Funding was provided by NSF grants OIA-0939564 to the
Center for Dark Energy Biosphere Investigations (JPA and subawards to JM and BNO), OCE-
1130146 (CGW), and OCE-1131210 and OCE-1260408 to Andy T. Fisher. This is C-DEBI
contribution number 430 and a DCO contribution.

101

Figure 1. Overview of the Dorado Outcrop (A), locations of sediment core samples (D), and
sediment geochemical profiles (B,C,E,F). A) The Dorado Outcrop is located in the Eastern
Tropical Pacific Ocean on a 23-million-year old ridge flank of the East Pacific Rise. B) Sediment
profiles of porefluid nitrate concentrations with depth (centimeters below seafloor, cmbsf). C)
Dissolved oxygen concentrations of sediment push cores from Wheat et al., 2017. D)
Bathymetric map of Dorado Outcrop, which is approximately 2 km long and 0.5 km wide, rising
approximately 150 m above the surrounding thickly sedimented seafloor. Push core locations are
marked by circles, with the corresponding core designation labeled next to the circles. Adapted
from Wheat et al., 2017. E) Dissolved and (F) solid phase manganese profiles from sediment
102
push cores. In B-E, and F, sediment cores were classified based on nitrate profiles, which have
been shown to demonstrate advective fluid flow from the ocean basement into the thin sediments
on Dorado Outcrop (Wheat and Fisher, 2008). Symbol color represents hydrothermal (red),
intermediate (blue), and background sediment (black) groups. Grey areas represent where
samples were collected for 16S rRNA gene analysis.

Figure 2. Schematic of the Cool Hydrothermal System (CHS) at Dorado Outcrop. Seawater
enters the crust through recharging outcrops, such as Tengosed, 20 km distance from Dorado
Outcrop. This cool seawater flows through the crust for < 3 years before discharging at Dorado
Outcrop (Wheat et al., 2017). The red circle denotes the focus of this study, the sediments on and
near Dorado which experience hydrothermal fluid flux through the porewater.

103

Figure 3. Taxonomic abundance of classes over 1% of the total community in Dorado Outcrop
sediment, basalt, and bottom seawater samples based on 16S rRNA gene sequencing using the
EMP primer pair (515F – 806R). Basalts and bottom seawater samples from Lee et al., 2015.
104

105
Figure 4. Select ASVs and their taxonomic assignments which were more abundant in the
hydrothermal and/or the hydrothermal plus intermediate samples. Bars are colored by sample,
and divided by ASV abundance Relative abundance is in percent of total sequences per sample.
The deepest taxonomic level assigned are listed next to the phylum followed by the number of
ASVs graphed.
106

107
Figure 5. Select ASVs and their taxonomic assignments which were more abundant in the
background and/or the background plus intermediate samples. Bars are colored by sample, and
divided by ASV abundance Relative abundance is in percent of total sequences per sample. The
deepest taxonomic level assigned are listed next to the phylum followed by the number of ASVs
graphed.
108

Figure 6. A) Principal coordinates analysis (PCoA) of log transformed OTU abundance data
showing clustering of samples by similarity using the Bray-Curtis dissimilarity index based on
109
rank abundance, with points (samples) colored by sample type. B) PCoA as in A, with color
coded by nitrate concentration. C) Canonical correspondence analysis (CCA) of samples based
on Bray-Curtis dissimilarity index using log-transformed OTU frequencies, nitrate
concentrations, and solid phase manganese concentrations. Labels show the core number (PC#),
D indicates a deep sample (9-10 cmbsf) and S indicates a shallow sample (3-4 cmbsf). D) PCoA
of sediments (this study) and basalts, basaltic biofilm, and seawater samples (Lee et al., 2015).

Figure 7. Relative abundance of genera in Clarion Clipperton Zone sediments and Dorado
hydrothermal sediments. Phylum is labeled to the left of the genus name along the y-axis.

0 - 5 cm b sf 3 - 4 cm b sf 9 - 1 0 cm b sf 8 - 1 0 cm b sf
Alphaproteobacteria | Magnetospira
Nitrosomonas
Nitrosospira
Deferribacteres | Caldithrix
G55
H16
Haliangium
Psychrobium
Sedimenticola
Nitrospinae | Nitrospina
Nitrospirae | Nitrospira
Omnitrophica | Candidatus Omnitrophus
Candidatus Scalindua
Pir4 lineage
Planctomyces
Rhodopirellula
SM1A02
Urania-1B-19
Z195MB87
Thaumarchaea | Candidatus Nitrosopumilus
PC1
CCZ
0-5 cm
CCZ
8-10 cm
PC2
PC5
PC6
PC8
PC1
PC2
PC5
PC6
PC8
0%
5%
10%
15%
R e l a t i ve
a b u n d a n ce
Planctomycetes
Gammaproteobacteria
Deltaproteobacteria
Betaproteobacteria
Clarion Clipperton Dorado Outcrop
110
Table 1. Characteristics of sediment samples from on and near Dorado Outcrop that were used
for 16S rRNA gene sequencing (see Figure 2 for more detail). Core designation based on
nitrate+nitrite concentrations, which were indicative of hydrothermal fluid flow through
sediments (Wheat et al., 2008). Oxygen values are from Wheat et al., 2017. A dash (-) indicates
that the value was not measured for this sample.
*
Nitrate+nitrite value from 2 cmbsf depth.
Core
Number
Core
Designation
Depth
(cmbsf)
Oxygen
(m M)
Nitrate+Nitrite
(m M)
Solid Phase Mn
(wt%)
PC1 Hydrothermal
3-4 - 43.01 0.818
9-10 - 37.31 1.252
PC2 Hydrothermal
3-4 5.9 34.68 1.324
9-10 - 39.02 0.534
PC3 Background
3-4 1.1 1.65 0.679
9-10 - 0.74 0.030
PC4 Intermediate
3-4 10.7 21.14
*
0.293
9-10 1.8 14.16 0.045
PC5 Hydrothermal
3-4 6.5 37.84 0.311
9-10 - 31.11 0.085
PC6 Hydrothermal
3-4 26.9 40.47 1.149
9-10 5.6 31.62 0.753
PC7 Intermediate
3-4 2.1 21.10 0.614
9-10 2.5 13.98 0.066
PC8 Hydrothermal
3-4 14.6 38.38 0.668
9-10 54.7 40.33 0.189
PC9 Intermediate
3-4 0.7 23.00 0.629
9-10 3.5 10.09 0.682

111
Supplemental Figures

Figure S1. Taxonomic assignment at the family level of sequences from extraction blanks.

112

Figure S2. Principal coordinate analysis (PCoA) based on weighted UniFrac analysis of all
sediment sample replicates (in red, blue, and black) before sourcetracker with
extraction/sequencing blanks (green). The closer the points are positioned to each other, the more
similar the samples are. The further away the samples are from another, the less similar the
samples are. Extraction/sequencing blanks are clustering away from all environmental samples.
113

Figure S3. Cluster diagram of all samples and replicates based on Bray-Curtis dissimilarity.
Replicates are color coded. Shallow samples are from 3-4 centimeters below seafloor. Deep
samples are from 9-10 centimeters below seafloor. Samples more closely positioned and sharing
nodes are more similar.
114

Figure S4. Class level assignments of major groups in sediments on Dorado Outcrop, based on
averages of triplicate samples.

115

Figure S5. Relative abundances of genera at Dorado Outcrop.
AKYG587
Anaerobacillus
Arcobacter
Blastopirellula
Caldithrix
Candidatus_Latescibacter
Candidatus_Nitrosopumilus
Candidatus_Omnitrophus
Candidatus_Paceibacter
Candidatus_Scalindua
Desulfatiglans
Desulfobacter
Desulfocurvus
Desulfovibrio
Dethiosulfatibacter
G55
H16
Haliangium
Haliea
Halioglobus
Magnetospira
Malonomonas
Marinobacter
Nitrosomonas
Nitrosospira
Nitrospina
Nitrospira
OM60(NOR5)_clade
PAUC26f
Pelagibius
Pelobacter
Peredibacter
Pir4_lineage
Planctomyces
Porticoccus
Pseudohongiella
Psychrobium
Ralstonia
Rhodopirellula
Sedimenticola
SEEP-SRB1
SM1A02
Spirochaeta_2
Sva0081_sediment_group
Thermomarinilinea
Tistlia
Ulvibacter
Urania-1B-19_marine_sediment_group
W4
Z195MB87
PC1S1
PC1S2
PC1S3
PC2S1
PC2S2
PC2S3
PC5S1
PC5S2
PC5S3
PC6S1
PC6S2
PC6S3
PC8S1
PC8S2
PC8S3
PC4S1
PC4S2
PC4S3
PC7S1
PC7S2
PC7S3
PC9S1
PC9S2
PC9S3
PC3S1
PC3S2
PC3S3
PC1D1
PC1D2
PC1D3
PC2D1
PC2D2
PC2D3
PC5D1
PC5D2
PC5D3
PC6D1
PC6D2
PC6D3
PC8D1
PC8D2
PC8D3
PC4D1
PC4D2
PC4D3
PC7D1
PC7D2
PC7D3
PC9D1
PC9D2
PC9D3
PC3D1
PC3D2
PC3D3
0.00
0.05
0.10
0.15
va l u e
Hydrothermal 3-4 cm Intermediate 3-4 cm BG 3-4 cm Hydrothermal 9-10 cm Intermediate 9-10 cm
BG 9-10
cm
116
117
Figure S6. Violin plots of percentages of specific taxa separated by sample type (B: background,
I: intermediate, H: hydrothermal) and depth (color coded as blue for 3-4 cmbsf and black for 9-
10 cmbsf). On the y-axis, the violin plots display the distribution of relative percentages of the
selected taxa in each sample type. Panels a – f show taxa whose relative percentages were
increased in hydrothermal sediments, relative to background or intermediate samples. Panels g –
l show taxa whose relative percentages were decreased in hydrothermal sediments.
118

Figure S7. The abundance of the top 100 ASVs and their taxonomic assignments which showed
no distinct pattern between sample types, for Acidobacteria and Actinobacteria (A);
119
Bacteroidetes, Chloroflexi, Gemmatimonadetes, Euryarcheaota, and Ignavibacteria (B);
Nitrospinae, Nitrospirae, and Planctomycetes (C); Alphaproteobacteria and Deltaproteobacteria
(D); Gammaproteobacteria (E); and Thaumarchaea (F). Relative abundance is in percent of total
sequences per sample. The taxonomic levels listed are phylum followed by the most specific
taxonomic level assigned (c = class, o = order, f = family, g = genus) and the taxon name.
OTU114 could not be assigned beyond the phylum level.

Figure S8. Combined relative abundance of the top 300 ASVs (as depicted in Figs. 4 and 5) as a
proportion of the total community. The top 300 ASVs represented over half of the sequences in
most samples, and included all ASVs over 1% abundance.

Table S1. Sequence statistics during processing in DADA2.
Sample core
Depth
(cmbsf)
Replicate
Input
Sequences
Filtered
Sequences
Denoised
Sequences
Non-chimeric
sequences
PC1

3-4
1 69,848 69,848 69,848 68,979
2
122,247 122,247 122,247 117,475
3 51,003 51,003 51,003 50,543
9-10
1 76,350 76,350 76,350 75,436
2 88,730 88,730 88,730 87,101
0.0
0.2
0.4
0.6
PC1S1
PC1S2
PC1S3
PC2S1
PC2S2
PC2S3
PC5S1
PC5S2
PC5S3
PC6S1
PC6S2
PC6S3
PC8S1
PC8S2
PC8S3
PC1D1
PC1D2
PC1D3
PC2D1
PC2D2
PC2D3
PC5D1
PC5D2
PC5D3
PC6D1
PC6D2
PC6D3
PC8D1
PC8D2
PC8D3
PC4S1
PC4S2
PC4S3
PC7S1
PC7S2
PC7S3
PC9S1
PC9S2
PC9S3
PC4D1
PC4D2
PC4D3
PC7D1
PC7D2
PC7D3
PC9D1
PC9D2
PC9D3
PC3S1
PC3S2
PC3S3
PC3D1
PC3D2
PC3D3
Relative Abundance
Total abundance of Top 300 OTUs, including all OTUs >1% total abundance
120
3
61,471 61,471 61,471 60,995
PC2

3-4
1
72,689 72,689 72,689 71,315
2
101,447 101,447 101,447 98,988
3
86,472 86,472 86,472 83,651
9-10
1
6,786 6,786 6,786 6,786
2
91,073 91,073 91,073 88,822
3
64,783 64,783 64,783 62,781
PC3

3-4
1
68,230 68,230 68,230 67,191
2
54,819 54,819 54,819 54,231
3
90,734 90,734 90,734 88,941
9-10
1
90,048 90,048 90,048 88,860
2
65,571 65,571 65,571 65,165
3
82,013 82,013 82,013 81,057
PC4
3-4
1
76,050 76,050 76,050 74,954
2
82,935 82,935 82,935 82,297
3
106,678 106,678 106,678 104,776
9-10
1
81,908 81,908 81,908 81,279
2
68,339 68,339 68,339 67,626
3
90,222 90,222 90,222 89,482
PC5

3-4
1
84,328 84,328 84,328 82,828
2
85,920 85,920 85,920 83,688
3
90,245 90,245 90,245 88,814
9-10
1
66,030 66,030 66,030 65,492
2
69,441 69,441 69,441 68,132
3
37,399 37,399 37,399 37,135
PC6

3-4
1
68,018 68,018 68,018 66,771
2
55,296 55,296 55,296 54,311
3
53,624 53,624 53,624 53,056
9-10
1
54,800 54,800 54,800 54,437
2
48,453 48,453 48,453 48,052
3
58,643 58,643 58,643 58,254
PC7

3-4
1
72,420 72,420 72,420 71,248
2
65,874 65,874 65,874 65,010
3
44,776 44,776 44,776 44,430
9-10
1
71,307 71,307 71,307 70,953
2
84,376 84,376 84,376 83,690
3
82,065 82,065 82,065 81,583
PC8

3-4
1
112,564 112,564 112,564 109,221
2
115,717 115,717 115,717 112,703
3
130,957 130,957 130,957 127,268
121
9-10
1
125,284 125,284 125,284 121,154
2
94,144 94,144 94,144 91,058
3
74,504 74,504 74,504 72,546
PC9

3-4
1
100,664 100,664 100,664 98,327
2
124,901 124,901 124,901 121,796
3
104,294 104,294 104,294 102,494
9-10
1
125,769 125,769 125,769 123,031
2
118,540 118,540 118,540 116,313
3
119,385 119,385 119,385 117,950
Extraction
blanks
-
1 98,527 90,263 90,263 50,937
2
73,836 68,294 68,294 42,542
3
96,769 88,320 88,320 55,279
Total
Sequences
4,659,316 4,637,061 4,637,061 4,459,234

Table S2. OTUs significantly correlated with nitrate or porefluid manganese concentrations.
Table S3. Proportion of samples assigned at the order level in samples.
Table S4. Taxonomic assignments and abundances of top 100 most abundant OTUs, which were
presenting in Figs. S5.
Table S5. Alpha diversity measures for each replicate.

122

“Oh, figures!' answered Ned. 'You can make figures do whatever you want.”
-Jules Verne, 20,000 Leagues Under the Sea

123
5. Microbial metabolism in low temperature hydrothermal sediments at the Dorado
Outcrop
In collaboration with Brandi Kiel Reese, James McManus, C. Geoffrey Wheat, Beth N. Orcutt,
and Jan P. Amend.
Abstract
Seamounts facilitate massive fluid fluxes through the marine lithosphere. Cool
Hydrothermal Systems (CHS) associated with these seamounts are likely widespread, but the
microbial communities associated with these systems have been rarely characterized. I analyzed
geochemistry and metagenomes from sediments associated with the Dorado Outcrop, a CHS in
the northeastern tropical Pacific. Geochemical conditions varied amongst sediment on and
adjacent to the outcrop versus nearby unaffected sediments. Similarly, microbial metabolic
potential varied between sediment types, with aerobic ammonium respiration and sulfur
compound oxidation genes more abundant in the oxic, nitrate-replete CHS-impacted cores than
in the background, anoxic sediments. Comparison of Dorado Outcrop metagenomes to those
from other deep sea sediments revealed that nitrogen metabolism and fermentation potential
were abundant in all sediment types, i.e. CHS, vent field sediments, and open ocean sediment.
However, Dorado Outcrop metagenomes lacked many of the hallmark hydrothermal taxa and
metabolisms, such as Epsilonproteobacteria and the reverse TriCarboxylic Acid (rTCA) cycle.
Instead, oxygen, nitrogen, and sulfur respiration characterized Dorado Outcrop metabolisms. I
conclude that the Dorado Outcrop sediment communities respond to geochemistry associated
with CHS and are similar to other low temperature deep sea sediment communities.

124
Introduction
Marine sediments harbor ~5 x 10
29
microorganisms, equivalent to the number of cells in
the global ocean, with diverse metabolisms that remain poorly understood (Kallmeyer et al.,
2012; Parkes et al., 2014). Depositional conditions, the initial seed communities, and diagenetic
reactions influence the structure and function of sediment microbial communities (Jorgensen,
2006; Walsh et al., 2016; Orsi et al., 2017). Deep sea sediments are historically understudied,
due largely to the difficulties associated with accessing these environments. The discovery of life
at hydrothermal vents in the 1970’s sparked interest in deep sea environments (Corliss et al.,
1979), including a large scientific focus on hot (upwards of 350°C degrees) hydrothermal vents
(Shanks, 2001). However, these vents are geographically limited and host specialized
communities, with only about 650 of these vents documented (E. T. Baker, 2017).
More recently, satellite gravimetry revealed more than 13,000 seamounts and outcrops
over 1.5 km in height across the seafloor (Wessel et al., 2010). Additionally, up to 25 million
smaller edifices, which are undetectable with current satellite technology, are hypothesized to
exist (Wessel et al., 2010). Based merely on these larger features, an estimated 10
14
kg yr
-1
of
mass flows through seamounts; a flux similar to that of mid-ocean ridges (Harris et al., 2004).
On the Cocos Plate in the eastern Pacific Ocean, a series of small outcrops facilitates regional
heat removal from the oceanic crust by allowing for fluid movement into and out of the shallow
lithosphere (Fisher et al., 2003). In this system, seawater enters the crust (i.e., recharges) through
some outcrops, undergoes slight but distinct geochemical changes in the crust, and discharges
through other nearby outcrops as cool (~12°C) hydrothermal fluid (Wheat and Fisher, 2008;
Wheat et al., 2017).
125
One confirmed CHS site is the Dorado Outcrop, which discharges approximately 20,000
L s
-1
of oxic (54.5 µM), nitrate-replete (38 µM) crustal fluid (Wheat and Fisher, 2008; Wheat et
al., 2017). In thin patches of sediment on and adjacent to the outcrop, fluid flux through the
sediment causes altered geochemical conditions relative to nearby unaffected sediments. In the
most CHS affected sediment pore fluids, nitrate concentrations are similar to those in seawater
(~40 µM) and remain consistent through the sediment core (Wheat and Fisher, 2008). Similarly,
input from the sediment-basalt interface causes oxygen concentrations to either remain elevated
through the sediment, or even increase with depth after a local minimum (Wheat et al., 2017).
Unaffected sediments nearby show typical diagenetic profiles, with geochemical indications of
anaerobic food web development.
Microbial community structure was recently linked with geochemical conditions in
systems where oxygen and nitrate flux upward from the basement into the overlying sediment
(Reese, Zinke, Sobol, Doug E LaRowe, et al., 2018). However, those geochemical gradients
were more gradual and on the orders of meters (Orcutt, Wheat, et al., 2013). At Dorado, the
oxygen and nitrate gradients occur over the submeter scale. How these conditions impact
microbial communities in this type of CHS sediment is not well understood.
Here, I use sediments collected from on and near the Dorado Outcrop to investigate: 1)
the metabolic potential of communities impacted by CHS-associated fluid flux, 2) the differences
between these communities relative to nearby “background” sediments, and 3) the relationship of
metabolisms to geochemistry. I then compare these communities to several other deep sea
marine sediments, both in active vent fields and underlying oligotrophic ocean provinces to
determine if the CHS sediments at Dorado are a unique ecosystem, or similar in metabolic and
taxonomic composition to other deep sea sediments.
126
Methods
Site description
The Dorado Outcrop is located between approximately 3,100 and 2,950 m water depth in
the Eastern Tropical Pacific Ocean (9°5’N, 87°5’W, Fig. 1A) (Wheat et al., 2017). Dorado
Outcrop is one of a series of outcrops which are thought to facilitate rapid flow of seawater
through the shallow lithosphere, where the entrained seawater is warmed and geochemically
altered, and then discharges through Dorado. Thin patches of sediment cover parts of the
outcrop, and the geochemical profiles of these sediments reflect crustal fluid migration from the
sediment-basalt interface (Wheat and Fisher, 2008).

Sample collection
Sediment samples were collected from the Dorado Outcrop from December 4-10, 2014
during cruise AT26-24 aboard the R/V Atlantis. Sediment push cores between and 10-28 cm was
retrieved during DSV Alvin dives 4780, 4782, and 4783. These cores will be referred to hereafter
as Push Core (PC) 3, 4, 6, and 8 (Figure 1B). Once shipboard, cores were examined for cracks or
seawater intrusion. Cores with no visible evidence of seawater intrusion were stored vertically at
4°C until pore fluid sampling or sediment sectioning began within 12 hours of being sampled.
For pore fluid sampling, Rhizons (Rhizosphere Research Products) were inserted through pre-
drilled side ports of individual sediment cores (e.g., (Ziebis, McManus, Ferdelman, Schmidt-
Schierhorn, Bach, Muratli, Edwards, and Villinger, 2012a). Dissolved oxygen data were
collected through side ports of companion cores and were published previously (Wheat et al.,
2017). Samples for bulk manganese compositions were determine from 1-cm to 10-cm intervals
that were selected at sea and refrigerated for further shore-based handling and analysis. For
biological sampling, sediment cores were sampled through the center of the core using sterile
127
plastic syringes that had been cut off at one end, wrapped in foil, and autoclaved before use. Core
sides were avoided to avoid possible seawater contamination. Multiple samples per depth were
taken from each sediment core in 6 cm intervals, and the samples were immediately frozen at -
70°C (the temperature of the shipboard freezers). Samples were shipped on dry ice to the
University of Southern California and were stored at -80°C until DNA extraction.
DNA extraction
DNA was extracted from the 9-10 cmbsf depth horizons of five sediment cores. Approximately 3
g of sediment was divided between four screw-cap 2-ml tubes. DNA was extracted in a UV-
sterilized clean hood using the FastDNA SPIN Kit (MP Biomedicals, Santa Ana, CA) using four
reactions for each replicate. The reactions for each sample were combined during the SPIN filter
step. DNA was eluted in 50 µL of molecular biology grade sterile water. Blank extraction
controls with no sample added were run alongside each extraction to verify sterility. Resulting
extractions were quantified using the Qubit HS dsDNA Assay on a Qubit 2.0 Fluorometer (Life
Technologies, Carlsbad, CA) following manufacturer protocols.
DNA Sequencing
A total of five samples were sequenced for metagenomics. Metagenomes from samples
Int_PC4 and HF_PC6a were sequenced at the Marine Biological Laboratories (Woods Hole,
MA, USA) thanks to the Deep Carbon Observatory’s Census of Deep Life. Metagenomic library
preparation and sequencing followed the Census of Deep Life protocol modified from Vineis et
al. using the Nextera DNA sample preparation kit following the manufacturer protocols
(Illumina, San Diego, CA, USA) (Vineis et al., 2016). The modifications to the Vineis protocol
were that the sequencing platform was the NextSeq (Illumina, San Diego, CA, USA), which
produced 150 base pair (bp) long paired end reads, and no microbiome enrichment step was
128
conducted. Samples HF_PC6b, BC_PC3, and HF_PC8 were sequenced at Molecular Research
DNA Lab (Shallowater, TX, USA). Libraries were prepared using Nextera DNA Sample
preparation kit following the manufacturer protocols with 25-50 ng DNA. Resulting libraries
were pooled and sequenced using the Illumina HiSeq platform to produce paired-end 150 bp
libraries.
Samples for 16S rRNA were sent to the Molecular Research DNA Lab for library
preparation and sequencing. The V4 region of the 16S rRNA gene was amplified using the Earth
Microbiome Project universal 515F (5ʹ-GTG CCA GCM GCC GCG GTA A-3’) and 806R (5ʹ-
GGA CTA CHV GGG TWT CTA AT-3’) primers (Caporaso et al., 2012). The forward primers
included 8 nucleotide barcodes. Libraries were created through PCR with HotStarTap Plus
(Qiagen, Germantown, MD) using the following protocol: 94°C for 3 minutes; 28 cycles of 94°C
for 30 seconds, 53°C for 40 seconds, and 72°C for 1 minute; and 72°C for 5 minutes. Amplified
DNA was pooled in approximately equimolar concentrations and purified using Ampure XP
beads (Beckman-Coulter, Indianapolis, IN). Amplified DNA was sequenced on an Illumina
MiSeq platform using Illumina TruSeq (Illumina, Inc., San Diego, CA) chemistry with 2 x 250
base pair chemistry.
Quality control and assembly
Metagenomics reads were trimmed using the program Trim Galore! v0.4.3 (Babraham
Bioinformatics, Cambridge, UK) in paired-end read mode, with a minimum quality score of 25, a
maximum 4 low quality bases before the read was trimmed, and a minimum post-trimming read
length of 80 bp. All Dorado metagenomes were co-assembled using Megahit v1.0.3-29-
g707d683 with a minimum contig size of 1000 bp (Li et al., 2015). Default kmer sizes of 21, 29,
129
39, 59, 79, 99, 119, and 141 were used for assembly. Contig names were simplified using anvi-
script-reformat-fasta in anvi’o v2.4.0 (Eren et al., 2015).
Read mapping and profile generation
Metagenomic reads were mapped to the assembled contigs using Bowtie2 v2.2.5 using
the ‘sensitive’ end-to-end setting. Resulting sam files were converted to bam files using samtools
v1.5, and these files were converted to anvi’o-compatible bam files in anvi’o. An anvi’o Anvi’o
database was created from the contigs, which included Open Reading Frame (ORF)
determination using Prodigal. Each sample was profiled against the contigs database using the
anvi’o command anvi-profile. A full project database was constructed from these profiles, and
included information about gene coverage and detection (percent over which the ORF was
mapped by reads). Tables with gene coverage by sample and gene detection by sample were
exported using the anvi’o command anvi-export-gene-coverage-and-detection.
Taxonomic and function assignments of genes
Protein coding sequences were exported from the contigs for taxonomic and functional
assignment using the anvi’o command anvi-get-aa-sequences-for-gene-calls. These sequences
were compared to the NCBI non-redundant database (accessed December, 2016) using the blastp
mode of DIAMOND v0.8.36 (Buchfink et al., 2014) with ‘sensitive’ settings and allowing only
one match per sequence. DIAMOND results were uploaded to MEGAN v6.10.2 and taxonomy
was assigned using the weighted Lowest Common Ancestor (LCA) assignment algorithm with a
minimum support percent identity of 0.3 (i.e. a taxon must have at least 0.3 percent, or 2,888
ORFs, assigned to be considered a ‘real’ hit). Taxonomic assignments for ORFs were exported
as a tab separated file and parsed in R.
130
Function was assigned to assembled genes using the GhostKoala server against the
KEGG database (accessed February, 2018). Functional assignments were exported as a tab
separated file and parsed in python and R. Pathway reconstruction was based on pathways in the
KEGG and Metacyc databases (Caspi et al., 2013), and published literature.
Coverage of coassembled genes was parsed from coverage profiles produced in anvi’o.
These were parsed by determining which genes contained more than 50% gene detection, and
averaging gene coverages by KO assignment. Abundances of reads were calculated by dividing
the number of reads per metagenomes were normalized to the abundance of recA reads in each
metagenome. Additionally, relative abundance percentages were calculated by dividing the
number of reads mapped to a gene by the total number of reads in the metagenome.
16S rRNA amplicon bioinformatics
Sequences from the 16S rRNA gene were trimmed of barcodes and low-quality
sequences using a quality cutoff of 25, and sequence read pairs were merged by MR DNA Lab,
resulting in an average sequence length of 299 base pairs. Merged sequences were processed in
Divisive Amplicon Denoising Algorithm 2 (DADA2) v1.6 following the protocol in Callahan et
al. (Ben J Callahan et al., 2016) and implemented in R version 3.4.1 (R Core Team, 2016). The
following analyses were performed in DADA2 unless otherwise stated. Sequence primers were
removed using Cutadapt (Marcel Martin, 2011). Sequence files containing all sequences were
split into individual files using the split_libraries_fastq.py and
split_sequence_file_on_sample_ids.py commands in QIIME (Caporaso, Kuczynski, Stombaugh,
Bittinger, Bushman, Costello, Fierer, Pe a, et al., 2010), including a minimum quality score
threshold of 25 for all sequences. These sequences were imported into DADA2, where they were
further filtered and trimmed to a length of 240 base pairs following the suggested DADA2
131
workflow. Core sequence variants, which are analogous to 100% sequence similarity Operational
Taxonomic Units (OTUs), were inferred from sequence data (Benjamin J Callahan et al., 2016).
The core sequence variants will be referred to as OTUs hereafter to avoid confusion. Sequences
representative of each OTU were chimera checked and chimeric OTUs were removed.
Sourcetracker was used to identify potential OTUs sourced from lab contamination or DNA
extraction kit contamination, and these OTUs were removed from samples before further
downstream analyses (Knights et al., 2011). Taxonomy (phylum through genus levels) was
assigned to OTUs using the Ribosomal Database Project’s naïve Bayesian classifier (Cole et al.,
2009) with the Silva v128 database as the reference (Pruesse et al., 2007). 16S rRNA sequences
were deposited into the National Center for Biotechnology Information (NCBI) Sequence Read
Archive (SRA) database under project PRJNA433243.
Comparison to other deep sea sediments
Samples from the Iheya vent fields (samples Iheya INT4, INT6, IRS6, and IRT2) as
described in Wang and Sun, 2017, were downloaded from NCBI SRA project PRNJA275476.
Samples from the Azores vent fields (Menez Gwen and Rainbow) as described in Cerquiera et
al. (2017) were downloaded from MG-RAST under project number mgp20053. The Iheya and
Azores samples were quality filtered as above, separately coassembled by publication (i.e., the
four Iheya samples were coassembled and Azores were coassembled), and were processed as
above.
Statistical analyses
All statistics were performed in R version 3.4.2. Ecological tests were performed using
the vegan package version 2.4-4 (Dixon, 2003). The ggplot2 package (Wickham 2009)
(Wickham, 2011) was used for graphing.
132
Results
Geochemistry
Concentration profiles of Total Organic Carbon (TOC), oxygen, nitrate, and ammonium
were different in sediment at Dorado Outcrop versus nearby background sediments (Fig 1C-F).
Cores PC6 and PC8 contained elevated concentrations of nitrate and oxygen and decreased TOC
relative to background sediments. Instead, the nitrate profiles were similar to Dorado Outcrop
sediments previously determined to be hydrothermally affected (referred hereafter as
hydrothermal fluid or HF), and core PC4 concentrations were intermediate between the
hydrothermal and background samples (Int) ((Wheat and Fisher, 2008)Zinke et al, in review). In
contrast, core PC3 from nearby sediment is considered a background sample (BG), and was
similar to the diffusion-limited sediments found nearby. These cores will be referred to as
HF_PC6 (note, two metagenomes were extracted and sequenced from PC6, and will be referred
to as HF_PC6a and HF_PC6b), HF_PC8, Int_PC4, and BG_PC3.
The HF cores contained overall less TOC than the BG and Int cores (Fig. 1C). In
HF_PC6, TOC was 1.32% at 0-2 cm depth, but decreased to 0.27% by 4-6 cm and remained less
than that concentration throughout the sediment core. HF_PC8 contained the least TOC, with
0.64 - 0.82% in the top 6 cm of sediment, and 0.12 - 0.19% in the 6-12 cm depth samples. At the
0-2 cm depth horizon, the intermediate core, Int_PC4 contained 1.41% TOC, but the TOC
decreased downcore to 0.2% by 23 cmbsf. TOC was greatest in the background sediment core,
BG_PC3, ranging between 1.08 - 1.49% of sediment dry weight.
Hydrothermal cores had measurable oxygen throughout the sediment column, ranging
between 2.1 - 60.2 µM oxygen through HF_PC8, with higher oxygen concentrations at the top
and bottom of the core (Fig. 1D). Oxygen concentrations decreased until 4 cmbsf before
133
increasing again close to the sediment-basement interface. HF_PC6 contained 5.6 - 30.1 µM
oxygen, with the highest oxygen concentrations at the 6 cmbsf and the lowest at 10 cmbsf. In
Int_PC4, oxygen decreased from 15.2 µM at 3 cm depth to 1.4 - 3.3 µM through the sediment
column, and increased again with depth to 7.2 µM at 24 cmbsf. The background core was
depleted in oxygen throughout the core relative to the HF cores, with oxygen concentrations
between 0 and 1.1 µM.
The nitrate concentrations in the hydrothermal cores (~30-40 µmol kg
-1
) were similar to
those measured in bottom seawater, and were elevated relative to the Int and BG nitrate
concentrations (Fig. 1E). Nitrate concentrations in the intermediate core (13.53 – 30.42 µmol kg
-
1
) were between those in the HF and BG cores. BG_PC3 had very low nitrate/nitrite
concentrations. The two HF cores contained the least and greatest ammonium concentrations,
ranging from less than 1 to 23.36 µM. The greatest ammonium concentrations in all cores were
within the top 10 cm. Ammonium concentrations in the pore fluids ranged from less than 1 µM -
4.29 µM in the BG_PC3 core and less than 1 - 9.51 µM in Int_PC4 (Fig 1F).

Sequencing and annotation
Metagenomic sequencing was performed on 5 samples from 9-10 cmbsf (two samples
from HF_PC6, referred to hereafter as HF_PC6a and HF_PC6b, HF_PC8, Int_PC4, BG_PC3).
This depth was chosen because concentrations of nitrate, oxygen, and ammonium varied between
these samples, and we previously found that this depth exhibited altered microbial communities
between hydrothermal, intermediate, and background sites (Zinke et al., in review). Sequencing
produced 213,324,150 high quality reads, totaling 37,453,729,610 base pairs (bp) (Table 1).
Assembly of these reads resulted in 521,954 contigs between 1000 and 169,639 bp long and
included 870,911,010 bp. Gene calling through Prodigal (Hyatt et al., 2010) and implemented in
134
anvi’o (Eren et al., 2015) predicted 1,214,256 genes coding sequences, of which 399,773 were
assigned to a KEGG Orthology (KO) (Kanehisa et al., 2016) function using GhostKoala.
Between 26.37% and 43.07% of the quality controlled metagenomes were mapped back to the
assembly using Bowtie2 in the sensitive mode (Langmead and Salzberg, 2012).
Gene sequences annotated with metabolic functions using GhostKoala were assigned
taxonomy using the Least Common Ancestor function in MEGAN (Huson, Beier, Flade, Górska,
El-Hadidi, Mitra, Ruscheweyh, and Tappu, 2016a) to identify: 1) the metabolic functions
possible in Dorado sediments and 2) the microbial populations that could be responsible for
biogeochemical transformations. Taxonomy is reported as percentage of the genes assigned
within a taxon relative to the total abundance of the gene. Gene abundances were determined by
mapping reads to reconstructed genes, which were normalized to the number of reads mapping to
the single copy DNA repair gene recombinase A (recA) in each sample, and are reported as this
ratio. It should be noted that activity cannot be inferred from metagenomes, and that data
discussed here represents genetic potential.
Microbial community taxonomy
Microbial community composition of sediments from 9-10 cmbsf were examined using
16S rRNA gene sequencing. Thaumarchaea, Planctomycetes, Alphaprotobacteria,
Deltaproteobacteria, Gammaproteobacteria, Chloroflexi, Nitrospirae, and Acidobacteria were the
main taxa identified in all four sites within the Dorado sediments based on 16S rRNA amplicon
sequencing (Fig. 2) (Zinke et al, in review). These phyla varied in relative abundance between
sample types. For example, Nitrospinae and Planctomycetes were most relatively abundant in the
BG sample compared to the Int and HF. Ignavibacteria were observed in a much greater relative
abundance in the BG sample than in the Int and HF samples, which contained < 1% of this
135
phylum. Candidatus Woesearchaeota were most abundant in the Int sample, and were less
abundant in the BG and HF samples. In contrast, Thaumarchaea, Alphaproteobacteria,
Gammaproteobacteria, and Nitrospira were most abundant in HF samples compared to Int and
BG samples.
Reconstructed 16S rRNA genes from the metagenomic coassembly were examined to
determine the presence of microbes without amplification primer biases. Bacterial 16S rRNA
genes (29 sequences, or 82.9% of recovered rRNA genes) and archaeal 16S rRNA (6 sequences,
or 17.1% of recovered rRNA genes) were recovered from the coassembly. At the phylum level,
the reconstructed sequences fell within the Deltaproteobacteria (17.1%), Acidobacteria (8.6%),
Actinobacteria (8.6%), Chloroflexi (8.6%), and Planctomycetes (5.7%). Four of the archaeal
sequences were most closely related to Thaumarchaea.
Nitrogen metabolisms
Numerous nitrogen metabolic pathways were abundant in all Dorado sediment samples,
with some metabolic pathways more abundant in HF metagenomes relative to the BG sample.
Nitrate reduction is encoded by nar and nitrite oxidation potential are encoded by nxr, both of
which are assigned within the same KEGG orthology and are genetically very similar
(Starkenburg et al., 2006). The nitrate transforming genes nar/nxr were the most abundant
respiratory genes (Fig. 3). The assigned taxonomies of these genes were similar between
samples, and included anammox bacteria within the Planctomycetes, Candidatus
Acetobacterium, Alphaproteobacteria, and Gammaproteobacteria (Fig. 2). In the HF samples,
some of these genes were assigned as Nitrospinia, and in the BG sediment only, a portion of the
subunits were assigned within Archaeoglobus. Genes for periplasmic nitrate reductase (napA)
were also present, though were much less abundant than nar/nxr (Fig. 3). The taxonomies that
136
were abundant for napA were Alphaproteobacteria, Deltaproteobacteria, Gammaproteobacteria,
and Nitrospinia.
DNRA-associated nitrite reductase gene nrfAH was present in Dorado sediments (0.05 -
0.08). In the HF and Int samples, nrfA was mostly assigned within Planctomycetes (Candidatus
Brocadia sinica) and Deltaproteobacteria (mostly within the order Myxococcales). The BG
sample had a large portion of nrfA assigned as Ignavibacteria (56.5%). All genes needed to
mediate denitrification were also found, including nirK or nirS, norBC, and nosZ. The nitrite
reductases, nirK and nirS, were assigned within anammox Planctomycetes (i.e. Candidatus
Scalindua brodae), Thaumarchaea, Gammaproteobacteria, and Betaproteobacteria. In the HF and
Int samples only, Nitrospinia were nirK genes were also detected. It should be noted that nitrite
reductases nirK and nirS also can facilitate the initial nitrite oxidation to NO in anammox
bacteria, such as Candidatus Scalindua brodae (van de Vossenberg et al., 2013). Hydrazine
synthase (hzs), a marker gene for anammox (Harhangi et al., 2012), was found in all samples in
abundances ranging between 0.02 - 0.08 recA normalized reads. The closest related hzs genes in
the NCBI database were from Candidatus Scalindua species. Assessing significant changes
based on small numbers of samples is difficult, but some trends in nitrogen compound reduction
potential emerged between samples. The HF and Int samples had increased normalized
abundances of nar relative to background, and the total of the denitrification pathway from
nitrite to N
2
was more abundant in HF samples than in the Int sample or BG sample.
Ammonium oxidation gene amo was most abundant in HF_PC8 compared to the other
samples (Fig. 3). Ammonia monooxygenase is very similar to particulate methane
monooxygenase (pmo), and are assigned to the same KEGG accession number. We manually
inspected taxonomic assignments and blasted these sequences against characterized organisms in
137
the NCBI database to determine if the bacterial amo/pmo genes were methane monooxygenases.
The amo/pmo genes were almost all most closely related to amo genes from known Ammonium
Oxidizing Archaea (AOA) or Ammonium Oxidizing Bacteria (AOB). These genes accounted for
100% of the amo/pmo genes recovered in the Int sample, 90.4 - 94.9% in the HF samples, and
34.8% in BG. Some of the genes in the BG sample were more likely pmo or were too divergent
to distinguish. In AOB, the gene hao encodes a hydroxylamine dehydrogenase, which converts
the hydroxylamine produced by ammonium monooxygenases into nitrite. These genes were
present and their taxonomy was assigned to known NOB (i.e. Nitrosomonas, Nitrosospira) and to
anammox bacteria. In total, HF samples contained more genetic potential for aerobic ammonium
oxidation than did the Int sample or the BG sample.
Sulfur Metabolisms
Genes involved in sulfur oxidation were generally more abundant in the HF samples
relative to the BG sample. Sox sulfur oxidation pathway genes were most abundant in HG
sediments and less abundant in the Int and BG samples. Sox sulfur oxidation system taxonomies
were assigned mostly within the Proteobacteria. Sulfite reductases encoded by soe genes have
been associated with the sox pathway (Dahl et al), and like the sox genes here, soe genes are
more abundant in the HF and Int sediment than the BG sample.
The dissimilatory sulfite reductase gene dsrAB was most abundant in the anoxic BG
sediments, but was found in the HF samples and Int sample as well. The dsrA gene was mostly
unable to be assigned at the phylum level, and no dsrA genes abundant in the Int sample were
assigned taxonomy. However, in the HF samples, 22.7 - 36.2% of dsrA reads were putatively
from Gammaproteobacteria. In the BG sample, Deltaproteobacteria (family Desulfobacteraceae)
138
reads were 15.2% of dsrA reads, with some reads assigned as Clostridia and the rest mostly
assigned as bacteria (60.6%), with no further taxonomic assignement available.
Sulfide:quinone oxidoreductase, encoded by sqr, reduces sulfide to S
0
was present in all
samples with no discrete trend between sample types (Fig. 3). Genes for sulfite reduction to
sulfate soe and asr were present in abundances ranging between 0.02 to 0.09 and < 0.01 to 0.03,
respectively. The asr sulfite reductases are anaerobic, and are more abundant in the BG sediment
than the HF and Int sediments (Fig. 3). Thiosulfate/polysulfide reductase genes (phr) were
present in relatively low abundance in all sediments, but was noticeably more abundant in the
BG sample relative to the HF and Int samples. Tetrathionate reductase genes (ttr) were present
only in the BG sample.
Carbon metabolism
Multiple fermentation pathways were found in Dorado sediment metagenomes, including
pathways that produce acetate, ethanol, propionate, and CO
2
. Alcohol (e.g. ethanol) production
was the most abundant fermentation pathway in all sediment types (Fig. 3). Genes encoding
acetate producing pathways and propionate producing pathways were also abundant, though less
so than ethanol pathways. Hydrogenases, which convert H
2
to 2H
+
and vice versa, were most
abundant in BG samples compared to Int and HF samples. Similarly, pyruvate formate lyase
(pfl), which converts pyruvate to formate and acetyl-CoA, was most abundant in the BG sample
and least abundant in the HF samples. Furthermore, formate dehydrogenase encoding gene fdh
was present in these sediments with at least 0.75 recA-normalized reads in each sample. Finally,
butyrate production potential and lactate production genes were present, but in lesser abundances
than ethanol, acetate, and propionate fermentation pathways.
139
Multiple genes considered markers of methane cycling (mmo (McDonald and Murrell,
1997), mxaF (Lau et al., 2013), and mcr (M. W. Friedrich, 2005) were not found, but genes
relating to methylamine and methanol utilization were. Genes associated with methylamine
consumption by aerobic bacteria (Gruffaz et al., 2014), such as methylamine-glutamate
methyltransferase (mgsABC), were prominent in HF and Int samples. The mau methylamine
dehydrogenase gene was found in approximately an order of magnitude greater abundance in HF
samples than in Int and BG samples. Dimethylamine/trimethylamine dehydrogenases (dmd/tmd)
were found in relatively high abundances in the HF samples as well. Methanol dehydrogenase
subunits (mxaCKL) were found in exclusively in HF and Int samples.
Methylamine metabolism encoding genes associated with anaerobic methylotrophy,
including methylamine corrinoid methyltransferases (mtb and mtm gene subunits) were most
abundant in the BG sediment, and less abundant in the Int and HF sediment. No methyl-
coenzyme M reductase genes (mcr) were assembled, so there is no evidence that these
methylamine utilization genes were involved with methane cycling. There were, however, genes
such as acsA, cdh subunits, and cooF, which are associated with the strictly anaerobic Wood-
Ljungdahl (WL) pathway. As with the methylamine corrinoid methyltransferases, these WL
genes were most abundant in the BG samples.
Carbon fixation pathways were also investigated in these sediments. Key genes in the
Calvin Benson Bassham (CBB) pathway are cbbL, cbbS, and prkB, which encode for RuBisCo
subunits (cbbLS) and phosphoribulokinase (prkB). These genes were found in samples here, and
in relatively similar abundances between samples. Additionally, genes for the 3-
hydroxypropionate/4-hydroxybutyrate carbon fixation pathway, which is found in
Thaumarchaea, were present in these sediments. The reverse TriCarboxylic Acid (rTCA) cycle
140
is another common carbon fixation pathway amongst microorganisms, but key genes encoding
citryl-CoA synthetase (ccs) and citryl-coA lyase (ccl) were not present in any of the
metagenomes here.
Genes encoding cytochrome C oxidase (cox) were notably abundant in Dorado
sediments, and were most abundant in the HF samples, relatively abundant in the Int sample, and
approximately half as abundant in the BG sample. The values for this aerobic respiration gene
are similar to the nitrate reductase (nar) abundances for HF and Int samples. In the BG sample,
however, cox is less abundant than nar. Most of the cox genes at Dorado were assigned as
subunits in coxABC, and were taxonomically assigned within many classes, including
Candidatus Tectomicrobia, candidate division NC10, Rhizobiales, Rhodospirillales,
Verrucomicrobiales.
Several reductive dehalogenating genes were found in sediments here. For instance, all
sediments contained haloalkane dehalogenase gene dhaA (Fig. 3). In the case of the HF samples,
dhaA was over an order of magnitude more relatively abundant than dsrAB. Taxonomic
assignment showed diverse community members contained this gene, including
Alphaproteobacteria, Deltaproteobacteria, Gammaproteobacteria, and Flavobacteriia.
Haloacetate dehalogenase gene dehH and 2-haloacid dehalogenase genes were less abundant but
still present. The dehalogenase encoding gene dehH was assigned mainly as Alphaproteobacteria
and Gammaproteobacteria in the HF samples. In the Int and BG sample, Betaproteobacteria are
more prevalent.
Comparison to other vent sites
Dorado Outcrop sediment communities were compared to other deep sea sediments near
hydrothermal activity at the Azores Vent Fields (Atlantic Ocean) and the Iheya Vent Fields
141
(Pacific Ocean) (Cerqueira et al., 2015; H.-L. Wang et al., 2017; Cerqueira et al., 2017). Several
similarities between the metagenomes from these three settings were apparent (Fig. 4). In all
sediments, genes encoding cytochrome C oxidases (cox genes) were amongst the most abundant
respiratory genes (sum of 1.22 - 4.39 recA normalized reads for all cox subunits). These genes
were most abundant in Dorado HF and Int sediments, and were similarly abundant in the Azores
Rainbow site. Nitrate reductases (nar and nap) were abundant in all sediments as well (1.07 -
4.53 recA normalized reads assigned to nar and nap subunits per sample). Compared to the
Azores and Iheya sediments, Dorado sediment (HF, Int, and BG) contained the most nitrate
reductase genes, and for all samples except Azores site Menez Gwen, nar genes were more
abundant than nap genes. The pathway for reduction of nitrite through denitrification was present
in all samples, but the abundance of denitrification genes was greatest in Dorado HF samples and
in the Azores Menez Gwen sample. DNRA was most abundant in the Dorado samples, found in
very low abundance in the Azores Rainbow sample, and not found in the rest of the samples
(Iheya samples and Menez Gwen). The Azores Rainbow sample was also the only sample with a
notable nitrogen fixation capability, and was the only sample similar to Dorado in the presence
of nitrification genes (i.e. amoABC and hao).
Genes related to sulfur compound oxidation (i.e. sox genes, soeAB, and sqr) were present
in all sites, but the Menez Gwen sediment contained by far the most sulfur compound oxidation
genes (9.21 recA normalized reads). The Rainbow sediments contained similar abundances of
sulfur compound oxidation genes as Dorado. Iheya sediments generally contained lower
abundances of sulfur oxidative potential. Sulfate reduction potential (based on dsrAB abundance)
vary between 0.02 and 1.05 recA normalized reads between all sediments. Menez Gwen and
142
Iheya samples INT4, INT6, and IRS6 were similar to Dorado samples in regards to dsrAB
abundance, but IRT2 and Rainbow contained greater dsrAB abundances.
Fermentation pathways were abundant in all samples, though ethanol and propionate
fermentation was less abundant in Menez Gwen and butyrate was more enriched relative to all
other samples. Reductive dehalogenation and methylamine utilization genes, which were present
in Dorado sediments, were also prevalent in all Iheya samples and in the Rainbow sample. The
rTCA cycle was not abundant in any sample except Menez Gwen, but the CBB and WL
pathways were present in most samples.
We further compared Dorado sediments to sediment metagenomes from the South Pacific
Gyre (Tully and Heidelberg, 2016). These sediments contained little TOC, and have previously
been shown to have nitrogen cycling capabilities. As at Dorado, aerobic respiration genes were
present, as were denitrification and DNRA genes, including narG, nirD, nirK, and norB. Genes
for ammonia monooxygenases, which facilitate nitrification, were present. Sulfur oxidizing,
sulfide oxidizing, some sox pathway genes, and dehalogenating genes were present in these
sediments as well.
Discussion
Dorado sediment biogeochemical cycling
CHS impact on sediments apparently supported aerobic and nitrate respiring
communities, whereas nearby background sediment reflected a typical anaerobic food web. For
instance, the migration of oxygen into the hydrothermal sediment column from below could
sustain aerobic respiration; correspondingly, oxygen respiration genes were most abundant in the
hydrothermal sediments. Aerobic respiration is often coupled to organic matter remineralization
in sediments, which usually proceeds most quickly under oxic conditions (Kristensen et al.,
143
1995). This likely explains why the TOC decreases most dramatically in the hydrothermal
samples, indicating that the input of oxygen due to crustal fluid impacts heterotrophic
metabolisms.
Similarly, the hydrothermal sediments contained relatively abundant aerobic ammonium
oxidation genes, which were assigned mostly within Thaumarchaea. The greatest abundance of
amo genes were found in HF_PC8, which correspondingly contained the most ammonium at the
~9 cm depth horizon. Aerobic ammonium oxidation in other sediments was dominated by
Thaumarchaea (Park et al., 2008; Dang et al., 2013), supporting the importance of AOA in
marine sediments globally.
Nitrate concentrations were also elevated in the hydrothermal sediments, and nitrogen
respiration genes were abundant (Fig 5). Genes encoding complete denitrification pathways and
potential for DNRA were present, providing a likely respiratory pathway for the microorganisms.
The oxidation of nitrite to nitrate was also likely in the HF sediments, as indicated by the
presence of nitrate/nitrite converting enzymes assigned to known NOB. Several of the genes
involved in denitrification and DNRA were assigned to known anammox bacteria, which are
widespread in marine sediments (Penton et al., 2006). Additionally, the key hydrazine forming
gene, hzs, was found. Anammox occurs under anaerobic conditions, and the HF cores were oxic,
so it unclear whether anammox was occurring in the HF sediments. However, in the intermediate
sample, nitrate and ammonium are present, and oxygen is depleted, so anammox was potentially
a key respiratory process here.
Considering the abundance of nitrogen cycling genes and the differences in both
geochemistry and genetic potential between sediment types here (i.e. HF, Int, and BG) it appears
that nitrogen respiration pathways were dominant in all sediments. Specifically, oxygen and
144
nitrate rich crustal fluid flux caused increased biological potential for aerobic ammonium
oxidation and nitrate reduction in the HF samples.
Sulfur cycling genes encoding both oxidation and reduction reactions were present in
Dorado sediment. The co-existence of genes encoding sulfur oxidizing and reducing enzymes
has previously been documented (Canfield et al., 2010). However, trends were apparent between
the oxygen and nitrate enriched HF samples versus the anaerobic BG sediments. One abundant
pathway in HF sediments was the sox pathway, which oxidizes a variety of reduced inorganic
sulfur compounds, such as thiosulfate (C. G. Friedrich et al., 2001). In contrast, many genes for
sulfur compound reduction were found in greater abundance or exclusively in the BG sample.
The canonical marker gene for sulfate reduction, dsr, was most abundant in the BG sediment,
indicating a switch from primarily sulfur compound oxidation to reduction between HF and BG
sediments. Some dsrA genes were found in low abundance in the HF samples, but most of these
genes were taxonomically assigned within sulfur oxidizing lineages, such as Thioalkalivibrio
(Lavy et al., 2017). Reverse dsr taxonomy generally reflects lineage phylogeny (Muller et al.,
2015), suggesting that the genes assigned within SOB here are reverse dsr genes. This
observation further indicates that sulfur compound oxidation is more dominant in HF sediments
and reduction is more dominant in sediments away from the outcrop.
Fermentation, small organic compound (e.g. methyamines, DMSO) utilization, and
reductive dehalogenation genes are widespread in marine sedimentary communities and
important drivers of sediment carbon cycling (Kirchman et al., 2014; Atashgahi et al., 2016;
Trembath-Reichert et al., 2017). Fermentation genes were abundant in all Dorado sediments, but
there were greater abundances of several methylamine utilization and reductive dehalogenation
genes in the HF and Int samples compared to the BG samples. This corresponded to the samples
145
which exhibited a more dramatic decrease in TOC with depth. This provides more evidence that
microbial remineralization potential is correlated with the increased oxidant (i.e. oxygen and
nitrate) availability, and that organic matter remineralization proceeds more quickly or
thoroughly in the oxic sediments, as previously has been demonstrated (Kristensen et al., 1995).
In contrast, the hydrogenase gene and Wood-Ljungdahl pathway abundances were highest in the
BG sample, reflecting the development of an anaerobic food web.
In short, it appears that microbial communities responded to geochemical conditions
associated with hydrothermal fluid migration at Dorado. Both Bacteria and Archaea, including
ammonium oxidizing archaea, were involved in biogeochemical cycling at Dorado. The flux of
oxygen and nitrate through HF sediments impacted the potential of microbial communities to
respire these compounds, correlated with the ability to oxidize sulfur and nitrogen compounds,
and increased the remineralization of organic matter in these sediments.
Deep sea sediments
The geochemistry, temperature, and vigor of hydrothermal vents affect local microbial
community composition (Campbell et al., 2013). The CHS at Dorado Outcrop is distinct from
previously characterized hydrothermal systems in its temperature, fluid chemistry, and age. As
such, Dorado microbial communities were expected to differ from recently characterized vent
sediments. We compared Dorado sediment metagenomes to metagenomes taken from the Iheya
Vent Fields in the western Pacific (H.-L. Wang et al., 2017) and the Azores Vent Fields from the
Atlantic (Cerqueira et al., 2015), which include vents significantly hotter than at Dorado, with
vent temperatures over 300°C measured in both vent fields. However, the most of sediments
investigated in these fields were distant from active vents and the sediments were cool (< 10°C).
146
Dorado HF samples were in many ways similar to these sediments - nitrogen and sulfur
cycling pathways were abundant, and include denitrification and sox pathways. Fermentation
was an abundant metabolic strategy in all sediments, and there was a noticeable lack of methyl-
CoM reductase in all sediments, pointing to a little anaerobic methane cycling in these low
temperature sediments, as found by Wang et al (2017) and Cerquiera et al (2017). These
similarities are likely explained by the in situ temperatures of sediments, which were around
3°C, and by the distance of these sediments from active vents, which 0.10 - 13.04 km from
sediment samples, implying that are under significant bottom seawater influence. However,
several major differences between Dorado and the cooler vent field samples were apparent,
including in the abundance of nitrification and DNRA genes at Dorado, and the presence of
sulfur oxidation genes such as soe at Dorado only.
The Azores Menez Gwen sediment community, which was taken only 4 m from an active
vent and was the warmest sediment (8.5°C) (Cerqueira et al., 2015), was most dissimilar to
Dorado HF samples (Fig. S1). This was evident in the abundance of Epsilonproteobacteria and of
key rTCA cycle genes, both of which are commonly found at hydrothermal vents (Shanks, 2001;
Campbell et al., 2006). This supports that Dorado microbial communities are not reflective of
‘hot’ hydrothermal communities, and that sediments in active vent fields but not impacted by
high temperatures are likely more reflective of background sediment conditions than of
hydrothermal regimes.
Deep sea sediments not associated with hydrothermal vents are undersampled, especially
in terms of metagenomics, but comparison between Dorado and several deep sea sites revealed
commonalities. Tully and Heidelberg sequenced metagenomes from South Pacific Gyre
sediments sampled at 0-5 cmbsf (2016). Reconstructed genomes from this site showed that
147
aerobic respiration, nitrate and nitrite reduction, and ammonium oxidation were key metabolic
strategies. Lineages capable of these processes were broadly similar to those found at Dorado,
including Nitrospira, Nitrospina, and Marine Group I Thaumarchaea (Tully and Heidelberg,
2016). . Further supporting this similarity, the reconstructed Thaumarchaea 16S rRNA sequences
were most closely related to Thaumarchaea from other cool deep sea sediments, such as South
Pacific Gyre sediment (Durbin and Teske, 2011). Additionally, organic lean sediments from the
Atlantic (North Pond) contained abundant nitrogen cycling microbes, and genes including narG,
nirBD, and nosZ assigned within Alphaproteobacteria and Betaproteobacteria were present
(Reese, Zinke, Sobol, Doug E LaRowe, et al., 2018).
Similarities between Dorado and deep sea sediment communities in sulfur cycling
potential exist. Sulfur oxidizing organisms in the East China Sea were assigned as
Alphaproteobacteria, Betaproteobacteria, and Gammaproteobacteria, including taxa found here
(e.g. Rhodospirilliaceae, Burkholderiaceae) (Yu Zhang et al., 2017). Through cultivation,
abundant thiosulfate-reducing Alphaproteobacteria were found in the sedimentary subseafloor
(Teske et al., 2000), which agrees with our metagenomics-based approach here.
Summary
Sediment communities on and nearby Dorado Outcrop were similar to other deep sea
sediments both taxonomically and functionally. However, key respiration processes were altered
by fluid migration associated with CHS venting. Nitrogen cycling potential was increased, most
dramatically in the potential for Thaumarchaea to oxidize ammonium. Sulfur respiration
pathways were altered, with oxidizing pathways more abundant in HF and reducing pathways
more abundant in BG sediments. In CHS, the heterotrophic remineralization of organic matter
148
was also apparently more thorough. Finally, Dorado was distinct from other sediments impacted
by hydrothermal systems, but with some strong similarities.
Acknowledgements
We would like to thank the entire Dorado Outcrop scientific party, the R/V Atlantis crew, and the
HOV Alvin crew on cruise AT26-24. We especially thank Annie Hartwell, Natalie Murray, Bo
Montagne, and Trevor Fournier for their assistance with fluid sampling, Mike Lee and Beate
Kraft for their assistance with sediment sampling, and Samuel Hulme and Chris Trabaol for
assistance with mapping. We thank Pratixa Savalia and Juan Orantes with laboratory assistance,
and Ben Tully, Elaine Graham, Mike Lee, and Lily Momper for help with bioinformatics. We
also thank the comments of two anonymous reviewers. Sequencing for this work was provided,
in part, by the Deep Carbon Observatory (DCO) Census of Deep Life, which is funded by the
Alfred P. Sloan Foundation. Funding was provided by NSF grants OIA-0939564 to the Center
for Dark Energy Biosphere Investigations (JPA and subawards to JM and BNO), OCE-1130146
(CGW), and OCE-1131210 and OCE-1260408 to Andy T. Fisher. This is a C-DEBI contribution
and a DCO contribution.

149

Figure 1. (A) Location of Dorado relative to nearby landmasses. (B) Bathymetric map of Dorado
150
with locations of sediment samples. (C) Total Organic Carbon (TOC) in weight percent; (D)
Oxygen concentrations, (E) nitrate+nitrite concentrations, and (F) ammonium concentrations
with respect to depth in sediment cores.

151

Figure 2. Taxonomic assignment of the 16S rRNA gene and select metabolic genes, represented
as percent of gene abundance assigned to each taxon.
152

Fig. 3. Metabolic gene abundances in Dorado Outcrop sediment. All genes abundances are
reported as ratio of number of genes mapped to the gene of interest divided by the number of
153
reads mapped to the recA gene.

Figure 4. Metabolic pathway abundance comparison between in Dorado Outcrop, Iheya Vent
154
Fields, and Azores Vent Fields samples. All genes abundances are reported as ratio of
number of genes mapped to the gene of interest divided by the number of reads mapped to
the recA gene.

Figure 5. Idealized schematic of biogeochemical cycling in hydrothermally influenced (HF;
average of the three hydrothermal metagenomes), intermediate (Int), and background (BG)
sediments at Dorado Outcrop. Circles next to gene names are scaled to gene abundances.
The average of the three HF samples are shown.

155
Table 1. Sequencing and assembly statistics

Sample
No. of raw
reads No. of raw BP
No. of trimmed
reads
PC8 62,923,411 9,318,189,320 57,345,531
PC6a 60,058,865 8,873,208,876 54,310,470
PC6b 25,233,840 3,762,923,743 19,119,602
PC4 65,203,557 9,789,212,364 44,637,299
PC3 39,506,681 5,710,195,307 37,911,248
Total 252,926,354 37,453,729,610 213,324,150
Coassembly statistics
No. of contigs 521,954
No. BP in
assembly 870,911,010
Avg Contig
Length (bp) 1,669
N50 1,592
Max contig
length (bp) 169,639
% mapping back Average: 34.24%
No. of CDS
*
1,214,256
No. KO
**

annotated 399,773
*CDS is coding sequence
**KO is KEGG orthologs

156
Supplemental Table 1. Phylum level (or class level for Proteobacteria) assignment of 16S rRNA
genes reconstructed in metagenomes
Phylum Number of reconstructed 16S rRNA
genes
Unknown 1
Acidobacteria 3
Actinobacterium 3
Alphaproteobacteria 1
Caldithrix 2
Candidatus Marinimicrobium 1
Candidatus Peregrinibacteria 1
Candidatus Woesebacteria 1
Chloroflexi 3
Deferribacterales 1
Deltaproteobacterium 6
Euryarchaea 2
Firmicutes 1
Gammaproteobacteria 1
Plactomycetes 2
Thaumarchaea 4

Figure S1. PCoA (Principal Coordinate Analysis) of metabolic gene abundance of the microbial
communities compared here.

157
“It was obvious that the matter had to be settled, and evasions were distasteful to me.”
-Jules Verne, 20,000 Leagues Under the Sea

158
6. Conclusions
The marine sedimentary biosphere covers over two thirds of Earth, but little of this
biosphere is explored. Here, I examined sediments from two systems to identify the microbial
communities and assess their relationship to geochemistry. These two sediments represent
contrasting endmembers – anoxic versus oxic, shallow sea versus deep sea, diffusion controlled
versus advection controlled. Both systems are opportunities to connect microbial communities to
geochemical regimes, i.e. methanogenic and aerobic, to further understand biogeochemistry in
the marine subsurface.
In the Baltic Sea, organic matter degradation dominated the metabolic landscape.
Extracellular carbohydrate active enzymes and peptidases were ubiquitous in sediments, and
significantly enriched in sediments corresponding to organic content. Some deviations from the
TOC and carbon degradation gene abundances could be attributed to marine influences during
deposition. Fermentation pathways were diverse, and some of these pathways could be correlated
to sediment organic content, total carbon content (organic content plus carbonate), and to
hydrology at the time of deposition. Furthermore, extracellular hydrolysis and fermentation were
active in at least two of the sediments, providing glimpses into the thriving deep biosphere in the
Baltic.
At the Dorado Outcrop, cool hydrothermal fluid flux through sediment provides the
microbial communities with replete high energy oxidants, i.e. oxygen and nitrate. This altered
the abundance of key microbial taxa in hydrothermal sediments relative to nearby anoxic
background sediments. Notably, ammonium oxidizing Thaumarchaea and nitrite oxidizing
Nitrospina and Nitrospira were enriched in the hydrothermal sediment, and anaerobic groups,
such as the Syntrophobacteriaceae were more abundant in the background sediment. Genes
159
involved in oxygen reduction, nitrate reduction, nitrite oxidation, sulfur oxidation, and
ammonium oxidation were also more abundant in these oxic, nitrate-rich hydrothermal
sediments. These gene abundances likely lead to more rapid organic matter mineralization, and
potentially cryptic nitrogen cycling in these sediments. Interestingly, anammox potential was
found in all sediments, regardless of oxygen and nitrogen contents. While the presence of these
genes do not equate to anammox activity in situ, but it might represent a community poised to
switch to anaerobic respiration strategies if hydrothermal activity turns quiescent.
Microbial communities in both systems were significantly correlated with to in situ
environment. In the Baltic, organic content was apparently the over-riding control on
heterotrophic potential. At Dorado Outcrop, nitrate and oxygen presence determined community
metabolism. However, in both of these systems, discrepancies between some available
metabolisms (based on gene presence) and geochemical parameters were apparent. Communities
from the Kattegat region, which is a marine entryway into the Baltic Sea, were enriched in
heterotrophic extracellular enzymes despite low organic content. Dorado Outcrop communities
from oxic and/or nitrate-rich sediment still had predicted genes for anammox. Together, these
results indicate that community metabolism is controlled by geochemistry, but that key
exceptions exist, and represent future avenues of subsurface exploration.

160
References
Ale, M.T., Mikkelsen, J.D., and Meyer, A.S. (2011) Important Determinants for Fucoidan
Bioactivity: A Critical Review of Structure-Function Relations and Extraction Methods for
Fucose-Containing Sulfated Polysaccharides from Brown Seaweeds. Marine Drugs 9:.
Aller, R.C. and Blair, N.E. (2004) Early diagenetic remineralization of sedimentary organic C in
the Gulf of Papua deltaic complex (Papua New Guinea): Net loss of terrestrial C and
diagenetic fractionation of C isotopes. Geochimica et Cosmochimica Acta 68: 1815–1825.
Alonso-Sáez, L., Waller, A.S., Mende, D.R., Bakker, K., Farnelid, H., Yager, P.L., et al. (2012)
Role for urea in nitrification by polar marine Archaea. PNAS 109: 17989–17994.
Alperin, M.J., Albert, D.B., and Martens, C.S. (1994) Seasonal variations in production and
consumption rates of dissolved organic carbon in an organic-rich coastal sediment.
Geochimica et Cosmochimica Acta 58: 4909–4930.
Amador-Noguez, D., Brasg, I.A., Feng, X.-J., Roquet, N., and Rabinowitz, J.D. (2011)
Metabolome Remodeling during the Acidogenic-Solventogenic Transition in Clostridium
acetobutylicum. AEM 77: 7984–7997.
Andersson, P.S., Wasserburg, G.J., and Ingri, J. (1992) The sources and transport of Sr and Nd
isotopes in the Baltic Sea. Earth and Planetary Science Letters 113: 459–472.
Andrén, T., Barker Jørgensen, B., Cotterill, C., Green, S., the IODP expedition 347 scientific
party (2015) IODP expedition 347: Baltic Sea basin paleoenvironment and biosphere.
Scientific Drilling 20: 1–12.
Andrén, T., Björck, S., Andrén, E., Conley, D., Zillén, L., and Anjar, J. (2011a) The
Development of the Baltic Sea Basin During the Last 130 ka. In, Harff,J., Björck,S., and
Hoth,P. (eds), The Baltic Sea Basin, The Baltic Sea Basin. Springer Berlin Heidelberg,
Berlin, Heidelberg, pp. 75–97.
Andrén, T., Björck, S., Andrén, E., Conley, D., Zillén, L., and Anjar, J. (2011b) The
Development of the Baltic Sea Basin During the Last 130 ka. In, Harff,J., Björck,S., and
Hoth,P. (eds), The Baltic Sea Basin, The Baltic Sea Basin. Springer Berlin Heidelberg,
Berlin, Heidelberg, pp. 75–97.
Andrén, T., Jorgensen, B.B., Cotterill, C., Green, S., Andrén, E., Ash, J., et al. (2015)
Proceedings of the IODP Andrén,T., Jørgensen,B.B., Cotterill,C., Green,S., Expedition 347
Scientists (eds) Integrated Ocean Drilling Program, College Station.
Arndt, S., Jørgensen, B.B., LaRowe, D.E., Middelburg, J.J., Pancost, R.D., and Regnier, P.
(2013) Quantifying the degradation of organic matter in marine sediments: A review and
synthesis. Earth-Science Reviews 123 IS -: 53–86.
Arnosti, C. (2011) Microbial Extracellular Enzymes and the Marine Carbon Cycle. Annu. Rev.
Mar. Sci. 3: 401–425.
Arnosti, C., Bell, C., Moorhead, D.L., Sinsabaugh, R.L., Steen, A.D., Stromberger, M., et al.
(2014) Extracellular enzymes in terrestrial, freshwater, and marine environments:
perspectives on system variability and common research needs. Biogeochemistry 117: 5–21.
Atashgahi, S., Lu, Y., and Smidt, H. (2016) Overview of Known Organohalide-Respiring
Bacteria—Phylogenetic Diversity and Environmental Distribution. In, Adrian,L. and
Löffler,F.E. (eds), Organohalide-Respiring Bacteria, Organohalide-Respiring Bacteria.
Springer Berlin Heidelberg, Berlin, Heidelberg, pp. 63–105.
Baker, B.J., Lazar, C.S., Teske, A.P., and Dick, G.J. (2015) Genomic resolution of linkages in
carbon, nitrogen, and sulfur cycling among widespread estuary sediment bacteria.
161
Microbiome 3: 14.
Baker, E.T. (2017) Exploring the ocean for hydrothermal venting: New techniques, new
discoveries, new insights. Ore Geology Reviews 86: 55–69.
Bartlett, D.H. (2002) Pressure effects on in vivo microbial processes. Biochimica et Biophysica
Acta (BBA) - Protein Structure and Molecular Enzymology 1595: 367–381.
Berg, R.D. and Solomon, E.A. Geochemical constraints on the distribution and rates of
debromination in the deep subseafloor biosphere. Geochimica et Cosmochimica Acta 174 IS
-: 30–41.
Berlemont, R. and Martiny, A.C. (2016) Glycoside Hydrolases across Environmental Microbial
Communities. PLoS Comput Biol 12: e1005300.
Beulig, F., Roy, H., Glombitza, C., and Jørgensen, B.B. (2018) Control on rate and pathway of
anaerobic organic carbon degradation in the seabed. PNAS 115: 367–372.
Biddle, J.F., Lipp, J.S., Lever, M.A., Lloyd, K.G., Sorensen, K.B., Anderson, R., et al. (2006)
Heterotrophic Archaea dominate sedimentary subsurface ecosystems off Peru. PNAS 103:
3846–3851.
Bodeï, S., Buatier, M., Steinmann, M., Adatte, T., and Wheat, C.G. (2008) Characterization of
metalliferous sediment from a low-temperature hydrothermal environment on the Eastern
Flank of the East Pacific Rise. Mar. Geol. 250: 128–141.
Boraston, A.B., Bolam, D.N., Gilbert, H.J., and Davies, G.J. (2004) Carbohydrate-binding
modules: fine-tuning polysaccharide recognition. Biochem. J. 382: 769–781.
Bowles, M.W., Mogollon, J.M., Kasten, S., Zabel, M., and Hinrichs, K.U. (2014) Global rates of
marine sulfate reduction and implications for sub-sea-floor metabolic activities. Science 344:
889–891.
Boyer, J.N. (1994) Aerobic and anaerobic degradation and mineralization of 14C-chitin by water
column and sediment inocula of the York River estuary, Virginia. AEM 60: 174–179.
Bray, J.R. and Curtis, J.T. (1957) An Ordination of the Upland Forest Communities of Southern
Wisconsin. Ecological Monographs 27: 325–349.
Buchfink, B., Xie, C., and Huson, D.H. (2014) Fast and sensitive protein alignment using
DIAMOND. Nat Meth 12: 59–60.
Buongiorno, J., Turner, S., Webster, G., Asai, M., Shumaker, A.K., Roy, T., et al. (2017) Inter-
laboratory quantification of Bacteria and Archaea in deeply buried sediments of the Baltic
Sea (IODP Expedition 347). FEMS Microbiology Ecology 93: fix007–fix007.
Callahan, Ben J, Sankaran, K., Fukuyama, J.A., McMurdie, P.J., and Holmes, S.P. (2016)
Bioconductor Workflow for Microbiome Data Analysis: from raw reads to community
analyses. F1000Res 5: 1492.
Callahan, Benjamin J, McMurdie, P.J., Rosen, M.J., Han, A.W., Johnson, A.J.A., and Holmes,
S.P. (2016) DADA2: High-resolution sample inference from Illumina amplicon data. Nat
Meth 13: 581–583.
Calvert, S.E. (1987) Oceanographic controls on the accumulation of organic matter in marine
sediments. Geological Society, London, Special Publications 26: 137–151.
Campbell, B.J., Engel, A.S., Porter, M.L., and Takai, K. (2006) The versatile ε-proteobacteria:
key players in sulphidic habitats. Nat Rev Micro 4: 458–468.
Campbell, B.J., Polson, S.W., Zeigler Allen, L., Williamson, S.J., Lee, C.K., Wommack, K.E.,
and Cary, S.C. (2013) Diffuse flow environments within basalt- and sediment-based
hydrothermal vent ecosystems harbor specialized microbial communities. Front. Microbiol.
4: 182.
162
Canfield, D.E., Stewart, F.J., Thamdrup, B., De Brabandere, L., Dalsgaard, T., DeLong, E.F., et
al. (2010) A Cryptic Sulfur Cycle in Oxygen-Minimum-Zone Waters off the Chilean Coast.
Science 330: 1375–1378.
Canfield, D.E., Thamdrup, B., and Hansen, J.W. (1993) The anaerobic degradation of organic
matter in Danish coastal sediments: Iron reduction, manganese reduction, and sulfate
reduction. Geochimica et Cosmochimica Acta 57: 3867–3883.
Canion, A., Overholt, W.A., Kostka, J.E., Huettel, M., Lavik, G., and Kuypers, M.M.M. (2014)
Temperature response of denitrification and anaerobic ammonium oxidation rates and
microbial community structure in Arctic fjord sediments. Environ Microbiol 16: 3331–3344.
Caporaso, J.G., Kuczynski, J., Stombaugh, J., Bittinger, K., Bushman, F.D., Costello, E.K.,
Fierer, N., Pe a, A.G., et al. (2010) QIIME allows analysis of high-throughput community
sequencing data. Nat Meth 7: 335–336.
Caporaso, J.G., Kuczynski, J., Stombaugh, J., Bittinger, K., Bushman, F.D., Costello, E.K.,
Fierer, N., Pena, A.G., et al. (2010) QIIME allows analysis of high-throughput community
sequencing data. Nat Meth 7: 335–336.
Caporaso, J.G., Lauber, C.L., Walters, W.A., Berg-Lyons, D., Huntley, J., Fierer, N., et al.
(2012) Ultra-high-throughput microbial community analysis on the Illumina HiSeq and
MiSeq platforms. ISME J 6: 1621–1624.
Carstensen, J., Andersen, J.H., Gustafsson, B.G., and Conley, D.J. (2014) Deoxygenation of the
Baltic Sea during the last century. PNAS 111: 5628–5633.
Caspi, R., Altman, T., Billington, R., Dreher, K., Foerster, H., Fulcher, C.A., et al. (2013) The
MetaCyc database of metabolic pathways and enzymes and the BioCyc collection of
Pathway/Genome Databases. Nucleic Acids Research 42: D459–D471.
Cerqueira, T., Pinho, D., Egas, C., Froufe, H., Altermark, B., Candeias, C., et al. (2015)
Microbial diversity in deep-sea sediments from the Menez Gwen hydrothermal vent system
of the Mid-Atlantic Ridge. Marine Genomics 24, Part 3 IS -: 343–355.
Cerqueira, T., Pinho, D., Froufe, H., Santos, R.S., Bettencourt, R., and Egas, C. (2017) Sediment
Microbial Diversity of Three Deep-Sea Hydrothermal Vents Southwest of the Azores.
Microbial Ecology.
Chen, X., Andersen, T.J., Morono, Y., Inagaki, F., Jorgensen, B.B., and Lever, M.A. (2017)
Bioturbation as a key driver behind the dominance of Bacteria over Archaea in near-surface
sediment. Scientific Reports 7: 2400.
Chung, I.K., Beardall, J., Mehta, S., Sahoo, D., and Stojkovic, S. (2011) Using marine
macroalgae for carbon sequestration: a critical appraisal. Journal of Applied Phycology 23:
877–886.
Cole, J.R., Wang, Q., Cardenas, E., Fish, J., Chai, B., Farris, R.J., et al. (2009) The Ribosomal
Database Project: improved alignments and new tools for rRNA analysis. Nucleic Acids
Research 37: D141–D145.
Conley, D.J., Björck, S., Bonsdorff, E., Carstensen, J., Destouni, G., Gustafsson, B.G., et al.
(2009) Hypoxia-Related Processes in the Baltic Sea. Environ. Sci. Technol. 43: 3412–3420.
Corinaldesi, C., Barucca, M., Luna, G.M., and Dell’Anno, A. (2011) Preservation, origin and
genetic imprint of extracellular DNA in permanently anoxic deep-sea sediments. Molecular
Ecology 20: 642–654.
Corliss, J.B., Dymond, J., Gordon, L.I., Edmond, J.M., Herzen, von, R.P., Ballard, R.D., et al.
(1979) Submarine Thermal Sprirngs on the Galapagos Rift. Science 203: 1073–1083.
D 'Hondt, S., Inagaki, F., Zarikian, C.A., Abrams, L.J., Dubois, N., Engelhardt, T., et al. (2015)
163
Presence of oxygen and aerobic communities from sea floor to basement in deep-sea
sediments. Nature Geosci 8: 299–304.
D'Hondt, S. (2004a) Distributions of Microbial Activities in Deep Subseafloor Sediments.
Science 306: 2216–2221.
D'Hondt, S. (2004b) Distributions of Microbial Activities in Deep Subseafloor Sediments.
Science 306: 2216–2221.
Dang, H., Luan, X., Zhao, J., and Li, J. (2009) Diverse and Novel nifH and nifH-Like Gene
Sequences in the Deep-Sea Methane Seep Sediments of the Okhotsk Sea. AEM 75: 2238–
2245.
Dang, H., Zhou, H., Yang, J., Ge, H., Jiao, N., Luan, X., et al. (2013) Thaumarchaeotal Signature
Gene Distribution in Sediments of the Northern South China Sea: an Indicator of the
Metabolic Intersection of the Marine Carbon, Nitrogen, and Phosphorus Cycles? AEM 79:
2137–2147.
Daniel, R., Stuertz, K., and Gottschalk, G. (1995) Biochemical and molecular characterization of
the oxidative branch of glycerol utilization by Citrobacter freundii. J Bacteriol 177: 4392–
4401.
Dauwe, B. and Middelburg, J.J. (2003) Amino acids and hexosamines as indicators of organic
matter degradation state in North Sea sediments. Limnol. Oceanogr. 43: 782–798.
Devol, A.H. (2015) Denitrification, Anammox, and N
2
Production in Marine Sediments. Annu.
Rev. Mar. Sci. 7: 403–423.
Dixon, P. (2003) VEGAN, a package of R functions for community ecology. Journal of
Vegetation Science 14: 927–930.
Dowd, S.E., Callaway, T.R., Wolcott, R.D., Sun, Y., McKeehan, T., Hagevoort, R.G., and
Edrington, T.S. (2008) Evaluation of the bacterial diversity in the feces of cattle using 16S
rDNA bacterial tag-encoded FLX amplicon pyrosequencing (bTEFAP). BMC Microbiology
8: 125.
Dowell, F., Cardman, Z., Dasarathy, S., Kellermann, M.Y., Lipp, J.S., Ruff, S.E., et al. (2016)
Microbial Communities in Methane- and Short Chain Alkane-Rich Hydrothermal Sediments
of Guaymas Basin. Front. Microbiol. 7: 17.
Dridi, B., Fardeau, M.-L., Ollivier, B., Raoult, D., and Drancourt, M. (2012)
Methanomassiliicoccus luminyensis gen. nov., sp. nov., a methanogenic archaeon isolated
from human faeces. IJSEM 62: 1902–1907.
Durbin, A.M. and Teske, A. (2011) Microbial diversity and stratification of South Pacific abyssal
marine sediments. Environ Microbiol 13: 3219–3234.
D’Hondt, S., Spivack, A.J., Pockalny, R., Ferdelman, T.G., Fischer, J.P., Kallmeyer, J., Abrams,
L.J., Smith, D.C., Graham, D., Hasiuk, F., Schrum, H., and Stancin, A.M. (2009a)
Subseafloor sedimentary life in the South Pacific Gyre. PNAS 106: 11651–11656.
D’Hondt, S., Spivack, A.J., Pockalny, R., Ferdelman, T.G., Fischer, J.P., Kallmeyer, J., Abrams,
L.J., Smith, D.C., Graham, D., Hasiuk, F., Schrum, H., and Stancin, A.M. (2009b)
Subseafloor sedimentary life in the South Pacific Gyre. PNAS 106: 11651–11656.
Edgar, R.C. (2010) Search and clustering orders of magnitude faster than BLAST.
Bioinformatics 26: 2460–2461.
Egger, M., Hagens, M., Sapart, C.J., Dijkstra, N., van Helmond, N.A.G.M., Mogollón, J.M., et
al. Iron oxide reduction in methane-rich deep Baltic Sea sediments. Geochimica et
Cosmochimica Acta 207 IS -: 256–276.
Egger, M., Hagens, M., Sapart, C.J., Dijkstra, N., van Helmond, N.A.G.M., Mogollón, J.M., et
164
al. (2017) Iron oxide reduction in methane-rich deep Baltic Sea sediments. Geochimica et
Cosmochimica Acta 207 IS -: 256–276.
Eren, A.M., Esen, Ö.C., Quince, C., Vineis, J.H., Morrison, H.G., Sogin, M.L., and Delmont,
T.O. (2015) Anvi’o: an advanced analysis and visualization platform for ‘omics data. PeerJ
3: e1319.
Finke, N. and Jorgensen, B.B. (2008) Response of fermentation and sulfate reduction to
experimental temperature changes in temperate and Arctic marine sediments. ISMEJ 2: 815–
829.
Finn, R.D., Bateman, A., Clements, J., Coggill, P., Eberhardt, R.Y., Eddy, S.R., et al. (2013)
Pfam: the protein families database. Nucleic Acids Research 42: D222–D230.
Fisher, A.T., Stein, C.A., Harris, R.N., Wang, K., Silver, E.A., Pfender, M., et al. (2003) Abrupt
thermal transition reveals hydrothermal boundary and role of seamounts within the Cocos
Plate. Geophys. Res. Lett. 30: n/a–n/a.
Flores, G.E., Campbell, J.H., Kirshtein, J.D., Meneghin, J., Podar, M., Steinberg, J.I., et al.
(2011) Microbial community structure of hydrothermal deposits from geochemically
different vent fields along the Mid-Atlantic Ridge. Environ Microbiol 13: 2158–2171.
Foght, J., Aislabie, J., Turner, S., Brown, C.E., Ryburn, J., Saul, D.J., and Lawson, W. (2004)
Culturable Bacteria in Subglacial Sediments and Ice from Two Southern Hemisphere
Glaciers. Microbial Ecology 47: 329–340.
Friberg, Y.B. The paleoceanography of Kattegat during the last deglaciation from benthic
foraminiferal stable isotopes.
Friedrich, C.G., Rother, D., Bardischewsky, F., Quentmeier, A., and Fischer, J. (2001) Oxidation
of Reduced Inorganic Sulfur Compounds by Bacteria: Emergence of a Common
Mechanism? AEM 67: 2873–2882.
Friedrich, M.W. (2005) Methyl-Coenzyme M Reductase Genes: Unique Functional Markers for
Methanogenic and Anaerobic Methane-Oxidizing Archaea. In, Methods in Methane
Metabolism, Part A, Methods in Enzymology. Academic Press, pp. 428–442.
Froelich, P.N., Klinkhammer, G.P., Bender, M.L., Luedtke, N.A., Heath, G.R., Cullen, D., et al.
(1979) Early oxidation of organic matter in pelagic sediments of the eastern equatorial
Atlantic: suboxic diagenesis. Geochimica et Cosmochimica Acta 43: 1075–1090.
Futagami, T., Morono, Y., Terada, T., Kaksonen, A.H., and Inagaki, F. (2009) Dehalogenation
Activities and Distribution of Reductive Dehalogenase Homologous Genes in Marine
Subsurface Sediments. AEM 75: 6905–6909.
Garcia, J.-L., Ollivier, B., and Whitman, W.B. (2006) The Order Methanomicrobiales. In, The
Prokaryotes. Springer New York, New York, NY, pp. 208–230.
Gilbert, J.A., Field, D., Huang, Y., Edwards, R., Li, W., Gilna, P., and Joint, I. (2008) Detection
of Large Numbers of Novel Sequences in the Metatranscriptomes of Complex Marine
Microbial Communities. PLOS ONE 3: e3042.
Girguis, P.R., Orphan, V.J., Hallam, S.J., and DeLong, E.F. (2003) Growth and Methane
Oxidation Rates of Anaerobic Methanotrophic Archaea in a Continuous-Flow Bioreactor.
AEM 69: 5472–5482.
Glombitza, C., Jaussi, M., Røy, H., Seidenkrantz, M.-S., Lomstein, B.A., and Jørgensen, B.B.
(2015) Formate, acetate, and propionate as substrates for sulfate reduction in sub-arctic
sediments of Southwest Greenland. Front. Microbiol. 6: 27.
Glombitza, C., Pedersen, J., Røy, H., and Jorgensen, B.B. (2014) Direct analysis of volatile fatty
acids in marine sediment porewater by two-dimensional ion chromatography-mass
165
spectrometry. Limnol. Oceanogr. Methods 12: 455–468.
Gooday, G.W. (1990) The Ecology of Chitin Degradation. In, Marshall,K.C. (ed), Advances in
Microbial Ecology, Advances in Microbial Ecology. Springer US, Boston, MA, pp. 387–
430.
Gosalbes, M.J., Durbán, A., Pignatelli, M., Abellan, J.J., Jiménez-Hernández, N., Pérez-Cobas,
A.E., et al. (2011) Metatranscriptomic Approach to Analyze the Functional Human Gut
Microbiota. PLOS ONE 6: e17447.
Gruffaz, C., Muller, E.E.L., Louhichi-Jelail, Y., Nelli, Y.R., Guichard, G., and Bringel, F. (2014)
Genes of the N-Methylglutamate Pathway Are Essential for Growth of Methylobacterium
extorquens DM4 with Monomethylamine. AEM 80: 3541–3550.
Haft, D.H., Selengut, J.D., Richter, R.A., Harkins, D., Basu, M.K., and Beck, E. (2012)
TIGRFAMs and Genome Properties in 2013. Nucleic Acids Research 41: D387–D395.
Hamdan, L.J., Coffin, R.B., Sikaroodi, M., Greinert, J., Treude, T., and Gillevet, P.M. (2012)
Ocean currents shape the microbiome of Arctic marine sediments. ISME J 7: 685–696.
Hans, M., Sievers, J., Müller, U., Bill, E., Vorholt, J.A., Linder, D., and Buckel, W. (1999) 2-
Hydroxyglutaryl-CoA dehydratase from Clostridium symbiosum. European Journal of
Biochemistry 265: 404–414 KW – 2–hydroxyglutaryl–CoA dehydratase.
Hardisty, D.S., Riedinger, N., Planavsky, N.J., Asael, D., Andrén, T., Jørgensen, B.B., and
Lyons, T.W. (2016) A Holocene history of dynamic water column redox conditions in the
Landsort Deep, Baltic Sea. Am J Sci 316: 713–745.
Harhangi, H.R., Le Roy, M., van Alen, T., Hu, B.-L., Groen, J., Kartal, B., et al. (2012)
Hydrazine Synthase, a Unique Phylomarker with Which To Study the Presence and
Biodiversity of Anammox Bacteria. AEM 78: 752–758.
Harris, R.N., Fisher, A.T., and Chapman, D.S. (2004) Fluid flow through seamounts and
implications for global mass fluxes. Geology 32: 725–728.
Häggblom, M.M. and Bossert, I.D. (2003) Halogenated Organic Compounds - A Global
Perspective. In, Häggblom,M.M. and Bossert,I.D. (eds), Dehalogenation: Microbial
Processes and Environmental Applications, Dehalogenation: Microbial Processes and
Environmental Applications. Springer US, Boston, MA, pp. 3–29.
Håvelsrud, O.E., Haverkamp, T.H., Kristensen, T., Jakobsen, K.S., and Rike, A.G. (2011) A
metagenomic study of methanotrophic microorganisms in Coal Oil Point seep sediments.
BMC Microbiology 11: 221.
Hedges, J.I. and Keil, R.G. (1995a) Sedimentary organic matter preservation: an assessment and
speculative synthesis. Marine Chemistry 49: 81–115.
Hedges, J.I. and Keil, R.G. (1995b) Sedimentary organic matter preservation: an assessment and
speculative synthesis. Marine Chemistry 49: 81–115.
Hedges, J.I., Keil, R.G., and Benner, R. (1997) What happens to terrestrial organic matter in the
ocean? Organic Geochemistry 27: 195–212.
Hoehler, T.M. and Jorgensen, B.B. (2013) Microbial life under extreme energy limitation. Nat
Rev Micro 11: 83–94.
Holmkvist, L., Ferdelman, T.G., and Jorgensen, B.B. A cryptic sulfur cycle driven by iron in the
methane zone of marine sediment (Aarhus Bay, Denmark). Geochimica et Cosmochimica
Acta 75: 3581–3599.
Hong, Y.-G., Li, M., Cao, H., and Gu, J.-D. (2011) Residence of Habitat-Specific Anammox
Bacteria in the Deep-Sea Subsurface Sediments of the South China Sea: Analyses of Marker
Gene Abundance with Physical Chemical Parameters. Microbial Ecology 62: 36–47.
166
Hood, E., Fellman, J., Spencer, R.G.M., Hernes, P.J., Edwards, R., D’Amore, D., and Scott, D.
(2009) Glaciers as a source of ancient and labile organic matter to the marine environment.
Nature 462: 1044 EP –.
Houmark-Nielsen, M. and Henrik Kjær, K. (2003) Southwest Scandinavia, 40–15 kyr BP:
palaeogeography and environmental change. J. Quaternary Sci. 18: 769–786.
Huber, J.A., Johnson, H.P., Butterfield, D.A., and Baross, J.A. (2006) Microbial life in ridge
flank crustal fluids. Environ Microbiol 8: 88–99.
Huson, D.H., Auch, A.F., Qi, J., and Schuster, S.C. (2007) MEGAN analysis of metagenomic
data. Genome Research 17: 377–386.
Huson, D.H., Beier, S., Flade, I., Górska, A., El-Hadidi, M., Mitra, S., Ruscheweyh, H.-J., and
Tappu, R. (2016a) MEGAN Community Edition - Interactive Exploration and Analysis of
Large-Scale Microbiome Sequencing Data. PLoS Comput Biol 12: e1004957.
Huson, D.H., Beier, S., Flade, I., Górska, A., El-Hadidi, M., Mitra, S., Ruscheweyh, H.-J., and
Tappu, R. (2016b) MEGAN Community Edition - Interactive Exploration and Analysis of
Large-Scale Microbiome Sequencing Data. PLoS Comput Biol 12: e1004957.
Huson, D.H., Mitra, S., Ruscheweyh, H.-J., Weber, N., and Schuster, S.C. (2011) Integrative
analysis of environmental sequences using MEGAN4. Genome Research 21: 1552–1560.
Hyatt, D., Chen, G.-L., LoCascio, P.F., Land, M.L., Larimer, F.W., and Hauser, L.J. (2010)
Prodigal: prokaryotic gene recognition and translation initiation site identification. BMC
Bioinformatics 11: 119.
Inagaki, F., Hinrichs, K.U., Kubo, Y., Bowles, M.W., Heuer, V.B., Hong, W.L., et al. (2015)
Exploring deep microbial life in coal-bearing sediment down to 2.5 km below the ocean
floor. Science 349: 420–424.
Inagaki, F., Nunoura, T., Nakagawa, S., Teske, A., Lever, M., Lauer, A., Suzuki, M., Takai, K.,
Delwiche, M., Colwell, F.S., Nealson, K.H., Horikoshi, K., D'Hondt, S., et al. (2006)
Biogeographical distribution and diversity of microbes in methane hydrate-bearing deep
marine sediments on the Pacific Ocean Margin. PNAS 103: 2815–2820.
Inagaki, F., Nunoura, T., Nakagawa, S., Teske, A., Lever, M., Lauer, A., Suzuki, M., Takai, K.,
Delwiche, M., Colwell, F.S., Nealson, K.H., Horikoshi, K., D’Hondt, S., et al. (2006)
Biogeographical distribution and diversity of microbes in methane hydrate-bearing deep
marine sediments on the Pacific Ocean Margin. PNAS 103: 2815–2820.
Inagaki, F., Suzuki, M., Takai, K., Oida, H., Sakamoto, T., Aoki, K., et al. (2003) Microbial
Communities Associated with Geological Horizons in Coastal Subseafloor Sediments from
the Sea of Okhotsk. AEM 69: 7224–7235.
Jansson, J.K., Neufeld, J.D., Moran, M.A., and Gilbert, J.A. (2012) Omics for understanding
microbial functional dynamics. Environ Microbiol 14: 1–3.
Jones, P., Binns, D., Chang, H.Y., Fraser, M., Li, W., McAnulla, C., et al. (2014) InterProScan 5:
genome-scale protein function classification. Bioinformatics 30: 1236–1240.
Jorgensen, B.B. (2006) Marine Geochemistry. In, Schulz,H.D. and Zabel,M. (eds), Bacteria and
Marine Biogeochemistry. Springer, Berlin, pp. 169–206.
Jorgensen, B.B. and Boetius, A. (2007) Feast and famine — microbial life in the deep-sea bed.
Nat Rev Micro 5: 770–781.
Jorgensen, B.B. and Marshall, I.P.G. (2016) Slow Microbial Life in the Seabed. Annu. Rev. Mar.
Sci. 8: 311–332.
Jormakka, M., Yokoyama, K., Yano, T., Tamakoshi, M., Akimoto, S., Shimamura, T., et al.
(2008) Molecular mechanism of energy conservation in polysulfide respiration. Nat Struct
167
Mol Biol 15: 730–737.
Jørgensen, S.L., Hannisdal, B., Lanzén, A., Baumberger, T., Flesland, K., Fonseca, R., et al.
(2012) Correlating microbial community profiles with geochemical data in highly stratified
sediments from the Arctic Mid-Ocean Ridge. PNAS 109: E2846–E2855.
Kallmeyer, J., Pockalny, R., Adhikari, R.R., Smith, D.C., and D’Hondt, S. (2012) Global
distribution of microbial abundance and biomass in subseafloor sediment. PNAS 109:
16213–16216.
Kanehisa, M., Sato, Y., Kawashima, M., Furumichi, M., and Tanabe, M. (2016) KEGG as a
reference resource for gene and protein annotation. Nucleic Acids Research 44: D457–D462.
Kirchman, D.L., Hanson, T.E., Cottrell, M.T., and Hamdan, L.J. (2014) Metagenomic analysis of
organic matter degradation in methane-rich Arctic Ocean sediments. Limnol. Oceanogr. 59:
548–559.
Kirkpatrick, J.B., Walsh, E.A., and D’Hondt, S. (2016) Fossil DNA persistence and decay in
marine sediment over hundred-thousand-year to million-year time scales. Geology 44: 615–
618.
Klippel, B., Sahm, K., Basner, A., Wiebusch, S., John, P., Lorenz, U., et al. (2014)
Carbohydrate-active enzymes identified by metagenomic analysis of deep-sea sediment
bacteria. Extremophiles 18: 853–863.
Knights, D., Kuczynski, J., Charlson, E.S., Zaneveld, J., Mozer, M.C., Collman, R.G., et al.
(2011) Bayesian community-wide culture-independent microbial source tracking. Nature
Methods 8: 761 EP –.
Kobayashi, T., Koide, O., Mori, K., Shimamura, S., Matsuura, T., Miura, T., et al. (2008)
Phylogenetic and enzymatic diversity of deep subseafloor aerobic microorganisms in
organics- and methane-rich sediments off Shimokita Peninsula. Extremophiles 12: 519–527.
Koonin, E.V., Makarova, K.S., and Aravind, L. (2001) Horizontal Gene Transfer in Prokaryotes:
Quantification and Classification. Annu. Rev. Microbiol. 55: 709–742.
Könneke, M., Bernhard, A.E., la Torre, de, J.R., Walker, C.B., Waterbury, J.B., and Stahl, D.A.
(2005) Isolation of an autotrophic ammonia-oxidizing marine archaeon. Nature 437: 543–
546.
Kristensen, E., Ahmed, S.I., and Devol, A.H. (1995) Aerobic and anaerobic decomposition of
organic matter in marine sediment: Which is fastest? Limnol. Oceanogr. 40: 1430–1437.
Kullenberg, G. and Jacobsen, T.S. (1981) The Baltic Sea: an outline of its physical
oceanography. Marine Pollution Bulletin 12: 183–186.
Langmead, B. and Salzberg, S.L. (2012) Fast gapped-read alignment with Bowtie 2. Nature
Methods 9: 357–359.
Lau, E., Fisher, M.C., Steudler, P.A., and Cavanaugh, C.M. (2013) The Methanol
Dehydrogenase Gene, mxaF, as a Functional and Phylogenetic Marker for Proteobacterial
Methanotrophs in Natural Environments. PLOS ONE 8: e56993.
Lavy, A., Keren, R., Yu, K., Thomas, B.C., Alvarez-Cohen, L., Banfield, J.F., and Ilan, M.
(2017) A novel Chromatiales bacterium is a potential sulfide oxidizer in multiple orders of
marine sponges. Environ Microbiol 20: 800–814.
Lee, M.D., Walworth, N.G., Sylvan, J.B., Edwards, K.J., and Orcutt, B.N. (2015) Microbial
Communities on Seafloor Basalts at Dorado Outcrop Reflect Level of Alteration and
Highlight Global Lithic Clades. Front. Microbiol. 6: 403.
Lever, M.A., Rouxel, O., Alt, J.C., Shimizu, N., Ono, S., Coggon, R.M., et al. (2013) Evidence
for Microbial Carbon and Sulfur Cycling in Deeply Buried Ridge Flank Basalt. Science 339:
168
1305–1308.
Li, D., Liu, C.-M., Luo, R., Sadakane, K., and Lam, T.-W. (2015) MEGAHIT: an ultra-fast
single-node solution for large and complex metagenomics assembly via succinct de Bruijn
graph. Bioinformatics 31: 1674–1676.
Li, D., Luo, R., Liu, C.-M., Leung, C.-M., Ting, H.-F., Sadakane, K., et al. (2016) MEGAHIT
v1.0: A fast and scalable metagenome assembler driven by advanced methodologies and
community practices. Pan-omics analysis of biological data 102: 3–11.
Lloyd, K.G., Alperin, M.J., and Teske, A. (2011) Environmental evidence for net methane
production and oxidation in putative ANaerobic MEthanotrophic (ANME) archaea. Environ
Microbiol 13: 2548–2564.
Lloyd, K.G., Lapham, L., and Teske, A. (2006a) An Anaerobic Methane-Oxidizing Community
of ANME-1b Archaea in Hypersaline Gulf of Mexico Sediments. AEM 72: 7218–7230.
Lloyd, K.G., Lapham, L., and Teske, A. (2006b) An Anaerobic Methane-Oxidizing Community
of ANME-1b Archaea in Hypersaline Gulf of Mexico Sediments. AEM 72: 7218–7230.
Lloyd, K.G., Schreiber, L., Petersen, D.G., Kjeldsen, K.U., Lever, M.A., Steen, A.D., et al.
(2013) Predominant archaea in marine sediments degrade detrital proteins. Nature 496: 215–
218.
Loffler, F.E., Yan, J., Ritalahti, K.M., Adrian, L., Edwards, E.A., Konstantinidis, K.T., et al.
(2013) Dehalococcoides mccartyi gen. nov., sp. nov., obligately organohalide-respiring
anaerobic bacteria relevant to halogen cycling and bioremediation, belong to a novel
bacterial class, Dehalococcoidia classis nov., order Dehalococcoidales ord. nov. and family
Dehalococcoidaceae fam. nov., within the phylum Chloroflexi. IJSEM 63: 625–635.
Lombard, V., Golaconda Ramulu, H., Drula, E., Coutinho, P.M., and Henrissat, B. (2013) The
carbohydrate-active enzymes database (CAZy) in 2013. Nucleic Acids Research 42: D490–
D495.
Lomstein, B.A., Langerhuus, A.T., D’Hondt, S., Jørgensen, B.B., and Spivack, A.J. (2012)
Endospore abundance, microbial growth and necromass turnover in deep sub-seafloor
sediment. Nature 484: 101–104.
Love, M.I., Huber, W., and Anders, S. (2014) Moderated estimation of fold change and
dispersion for RNA-seq data with DESeq2. Genome Biol. 15: 550.
Loy, A., Duller, S., Baranyi, C., Mußmann, M., Ott, J., Sharon, I., et al. (2009) Reverse
dissimilatory sulfite reductase as phylogenetic marker for a subgroup of sulfur-oxidizing
prokaryotes. Environ Microbiol 11: 289–299.
Lücker, S., Nowka, B., Rattei, T., Spieck, E., and Daims, H. (2013a) The Genome of Nitrospina
gracilis Illuminates the Metabolism and Evolution of the Major Marine Nitrite Oxidizer.
Front. Microbiol. 4: 27.
Lücker, S., Nowka, B., Rattei, T., Spieck, E., and Daims, H. (2013b) The Genome of Nitrospina
gracilis Illuminates the Metabolism and Evolution of the Major Marine Nitrite Oxidizer.
Front. Microbiol. 4: 27.
Lücker, S., Wagner, M., Maixner, F., Pelletier, E., Koch, H., Vacherie, B., et al. (2010) A
Nitrospira metagenome illuminates the physiology and evolution of globally important
nitrite-oxidizing bacteria. PNAS 107: 13479–13484.
Mackenzie, F.T., Lerman, A., and Andersson, A.J. (2004) Past and present of sediment and
carbon biogeochemical cycling models. Biogeosciences Discussions 1: 27–85.
Magoč, T. and Salzberg, S.L. (2011) FLASH: fast length adjustment of short reads to improve
genome assemblies. Bioinformatics 27: 2957–2963.
169
Manheim, F.T. (1966) A hydraulic squeezer for obtaining interstitial water from consolidated
and unconsolidated sediments. 550: 171–174.
Marshall, I.P.G., Karst, S.M., Nielsen, P.H., and Jorgensen, B.B. (2017) Metagenomes from deep
Baltic Sea sediments reveal how past and present environmental conditions determine
microbial community composition. Marine Genomics.
Marshall, I.P.G., Starnawski, P., Cupit, C., Fernández Cáceres, E., Ettema, T.J.G., Schramm, A.,
and Kjeldsen, K.U. (2017) The novel bacterial phylum Calditrichaeota is diverse,
widespread and abundant in marine sediments and has the capacity to degrade detrital
proteins. Environmental Microbiology Reports 9: 397–403.
Martin, Machovič and Štefan, J. (2008) Domain evolution in the GH13 pullulanase subfamily
with focus on the carbohydrate-binding module family 48. biolog 63: 1057.
Martin, Marcel (2011) Cutadapt removes adapter sequences from high-throughput sequencing
reads. EMBnet.journal; Vol 17, No 1: Next Generation Sequencing Data Analysis.
Martiny, A.C., Treseder, K., and Pusch, G. (2013) Phylogenetic conservatism of functional traits
in microorganisms. ISMEJ 7: 830 EP –.
Mason, O.U., Hazen, T.C., Borglin, S., Chain, P.S.G., Dubinsky, E.A., Fortney, J.L., et al.
(2012) Metagenome, metatranscriptome and single-cell sequencing reveal microbial
response to Deepwater Horizon oil spill. ISME J 6: 1715–1727.
Matos, M.N., Lozada, M., Anselmino, L.E., Musumeci, M.A., Henrissat, B., Jansson, J.K., et al.
(2016) Metagenomics unveils the attributes of the alginolytic guilds of sediments from four
distant cold coastal environments. Environ Microbiol 18: 4471–4484.
McDonald, I.R. and Murrell, J.C. (1997) The particulate methane monooxygenase gene pmoA
and its use as a functional gene probe for methanotrophs. FEMS Microbiology Letters 156:
205–210.
McMurdie, P.J. and Holmes, S. (2013) phyloseq: An R Package for Reproducible Interactive
Analysis and Graphics of Microbiome Census Data. PLOS ONE 8: e61217.
Middelburg, J.J., Vlug, T., Jaco, F., and van der Nat, W.A. (1993) Organic matter mineralization
in marine systems. Global and Planetary Change 8: 47–58.
Moyer, C.L., Dobbs, F.C., and Karl, D.M. (1995) Phylogenetic diversity of the bacterial
community from a microbial mat at an active, hydrothermal vent system, Loihi Seamount,
Hawaii. AEM 61: 1555–1562.
Muller, A.L., Kjeldsen, K.U., Rattei, T., Pester, M., and Loy, A. (2015) Phylogenetic and
environmental diversity of DsrAB-type dissimilatory (bi)sulfite reductases. ISME J 9: 1152–
1165.
Müller, J.A., Rosner, B.M., Abendroth, von, G., Meshulam-Simon, G., McCarty, P.L., and
Spormann, A.M. (2004) Molecular Identification of the Catabolic Vinyl Chloride Reductase
from Dehalococcoides sp. Strain VS and Its Environmental Distribution. AEM 70: 4880–
4888.
Nealson, K.H. (1997) Sediment Bacteria: Who“s There, What Are They Doing, and What”s
New? Annu. Rev. Earth Planet. Sci. 25: 403–434.
Notenboom, V., Boraston, A.B., Kilburn, D.G., and Rose, D.R. (2001) Crystal Structures of the
Family 9 Carbohydrate-Binding Module from Thermotoga maritimaXylanase 10A in Native
and Ligand-Bound Forms †,‡. Biochemistry 40: 6248–6256.
Orcutt, B.N., LaRowe, D.E., Biddle, J.F., Colwell, F.S., Glazer, B.T., Reese, B.K., et al. (2013)
Microbial activity in the marine deep biosphere: progress and prospects. Front. Microbiol. 4:
189.
170
Orcutt, B.N., Wheat, C.G., Rouxel, O., Hulme, S., Edwards, K.J., and Bach, W. (2013) Oxygen
consumption rates in subseafloor basaltic crust derived from a reaction transport model JA .
4: 2539 EP –.
Orsi, W.D., Coolen, M.J.L., Wuchter, C., He, L., More, K.D., Irigoien, X., et al. (2017) Climate
oscillations reflected within the microbiome of Arabian Sea sediments. Scientific Reports 7:
6040.
Orsi, W.D., Edgcomb, V.P., Christman, G.D., and Biddle, J.F. (2013) Gene expression in the
deep biosphere. Nature 499: 205–208.
Orsi, W.D., Richards, T.A., and Francis, W.R. (2018) Predicted microbial secretomes and their
target substrates in marine sediment. Nature Microbiology 3: 32–37.
Overbeek, R., Begley, T., Butler, R.M., Choudhuri, J.V., Chuang, H.-Y., Cohoon, M., et al.
(2005) The Subsystems Approach to Genome Annotation and its Use in the Project to
Annotate 1000 Genomes. Nucleic Acids Research 33: 5691–5702.
Pachiadaki, M.G., Rédou, V., Beaudoin, D.J., Burgaud, G., and Edgcomb, V.P. (2016) Fungal
and Prokaryotic Activities in the Marine Subsurface Biosphere at Peru Margin and
Canterbury Basin Inferred from RNA-Based Analyses and Microscopy. Front. Microbiol. 7:
364.
Park, S.-J., Park, B.-J., and Rhee, S.-K. (2008) Comparative analysis of archaeal 16S rRNA and
amoA genes to estimate the abundance and diversity of ammonia-oxidizing archaea in
marine sediments. Extremophiles 12: 605.
Parkes, R.J., Cragg, B., Roussel, E., Webster, G., Weightman, A., and Sass, H. (2014) A review
of prokaryotic populations and processes in sub-seafloor sediments, including
biosphere:geosphere interactions. Mar. Geol. 352 IS -: 409–425.
Parkes, R.J., Cragg, B.A., and Wellsbury, P. (2000) Recent studies on bacterial populations and
processes in subseafloor sediments: A review. Hydrogeology Journal 8: 11–28.
Parkes, R.J., Cragg, B.A., Bale, S.J., Getlifff, J.M., Goodman, K., Rochelle, P.A., et al. (1994)
Deep bacterial biosphere in Pacific Ocean sediments. Nature 371: 410–413.
Penton, C.R., Devol, A.H., and Tiedje, J.M. (2006) Molecular Evidence for the Broad
Distribution of Anaerobic Ammonium-Oxidizing Bacteria in Freshwater and Marine
Sediments. AEM 72: 6829–6832.
Petersen, K.R., Streett, D.A., Gerritsen, A.T., Hunter, S.S., and Settles, M.L. (2015) Super
deduper, fast PCR duplicate detection in fastq files ACM, Atlanta, GA.
Petersen, T.N., Brunak, S., Heijne, von, G., and Nielsen, H. SignalP 4.0: discriminating signal
peptides from transmembrane regions. Nature Methods 8: 785 EP –.
Poulicek, M. and Jeuniaux, C. (1991) Chitin biodegradation in marine environments: An
experimental approach. Biochemical Systematics and Ecology 19: 385–394.
Prosser, J.I., Head, I.M., and Stein, L.Y. (2014) The Family Nitrosomonadaceae. In,
Rosenberg,E., DeLong,E.F., Lory,S., Stackebrandt,E., and Thompson,F. (eds), The
Prokaryotes: Alphaproteobacteria and Betaproteobacteria. Springer Berlin Heidelberg,
Berlin, Heidelberg, pp. 901–918.
Pruesse, E., Quast, C., Knittel, K., Fuchs, B.M., Ludwig, W., Peplies, J., and Glockner, F.O.
(2007) SILVA: a comprehensive online resource for quality checked and aligned ribosomal
RNA sequence data compatible with ARB. Nucleic Acids Research 35: 7188–7196.
Quast, C., Pruesse, E., Yilmaz, P., Gerken, J., Schweer, T., Yarza, P., et al. (2012) The SILVA
ribosomal RNA gene database project: improved data processing and web-based tools.
Nucleic Acids Research 41: D590–D596.
171
R Core Team, R.C. (2016) R: A language and environment for statistical computing.
Raskin, L., Stromley, J.M., Rittmann, B.E., and Stahl, D.A. (1994) Group-specific 16S rRNA
hybridization probes to describe natural communities of methanogens. AEM 60: 1232–1240.
Rawlings, N.D. and Barrett, A.J. (1993) Evolutionary families of peptidases. Biochem. J. 290:
205–218.
Reed, D.W., Fujita, Y., Delwiche, M.E., Blackwelder, D.B., Sheridan, P.P., Uchida, T., and
Colwell, F.S. (2002) Microbial Communities from Methane Hydrate-Bearing Deep Marine
Sediments in a Forearc Basin. AEM 68: 3759–3770.
Reese, B.K., Mills, H.J., Dowd, S.E., and Morse, J.W. (2013) Linking Molecular Microbial
Ecology to Geochemistry in a Coastal Hypoxic Zone. Geomicrobiology Journal 30: 160–
172.
Reese, B.K., Zinke, L., Sobol, M.S., LaRowe, Douglas E, Zhang, X.-X., Jaekel, U., et al. (2018)
Nitrogen cycling of active bacteria within oligotrophic sediment of the Mid-Atlantic Ridge
flank. Geomicrobiology Journal.
Reese, B.K., Zinke, L.A., Sobol, M.S., LaRowe, Doug E, Orcutt, B.N., Zhang, X., et al. (2018)
Nitrogen Cycling of Active Bacteria within Oligotrophic Sediment of the Mid-Atlantic
Ridge Flank. Geomicrobiology Journal 62: 1–16.
Reyes, C., Schneider, D., Lipka, M., Thürmer, A., Böttcher, M.E., and Friedrich, M.W. (2017)
Nitrogen Metabolism Genes from Temperate Marine Sediments. Marine Biotechnology 19:
175–190.
Robador, A., Jungbluth, S.P., LaRowe, D.E., Bowers, R.M., Rappé, M.S., Amend, J.P., and
Cowen, J.P. (2014) Activity and phylogenetic diversity of sulfate-reducing microorganisms
in low-temperature subsurface fluids within the upper oceanic crust. Front. Microbiol. 5:
748.
Rosenberg, E., DeLong, E.F., Thompson, F.L., Lory, S., and Stackbrandt, E. (2014) The
Prokaryotes: Alphaproteobacteria and Betaproteobacteria Springer Berlin Heidelberg,
Berlin, Heidelberg.
Roy, H., Kallmeyer, J., Adhikari, R.R., Pockalny, R., Jørgensen, B.B., and D'Hondt, S. (2012)
Aerobic Microbial Respiration in 86-Million-Year-Old Deep-Sea Red Clay. Science 336:
922–925.
Röling, W.F.M., Milner, M.G., Jones, D.M., Lee, K., Daniel, F., Swannell, R.J.P., and Head,
I.M. (2002) Robust Hydrocarbon Degradation and Dynamics of Bacterial Communities
during Nutrient-Enhanced Oil Spill Bioremediation. AEM 68: 5537–5548.
Røy, H., Weber, H.S., Tarpgaard, I.H., Ferdelman, T.G., and Jørgensen, B.B. (2014)
Determination of dissimilatory sulfate reduction rates in marine sediment via radioactive 35S
tracer. Limnol. Oceanogr. Methods 12: 196–211.
Sansone, F.J. and Martens, C.S. (1982) Volatile fatty acid cycling in organic-rich marine
sediments. Geochimica et Cosmochimica Acta 46: 1575–1589.
Schippers, A., Neretin, L.N., Kallmeyer, J., Ferdelman, T.G., Cragg, B.A., John Parkes, R., and
Jørgensen, B.B. Prokaryotic cells of the deep sub-seafloor biosphere identified as living
bacteria. Nature 433: 861 EP –.
Schmidt, J.M. (2016) Microbial Extracellular Enzymes in Marine Sediments: Methods
Development and Potential Activities in the Baltic Sea Deep Biosphere. Masters Thesis.
Schweiger, G., Dutscho, R., and Buckel, W. (1987) Purification of 2-hydroxyglutaryl-CoA
dehydratase from Acidaminococcus fermentans. European Journal of Biochemistry 169:
441–448.
172
Seeberg-Elverfeldt, J., Schlüter, M., Feseker, T., and Kölling, M. (2005) Rhizon sampling of
pore waters near the sediment/water interface of aquatic systems. Limnol. Oceanogr.
Methods 3: 361–371.
Seitz, K.W., Lazar, C.S., Hinrichs, K.-U., Teske, A.P., and Baker, B.J. (2015) Genomic
reconstruction of a novel, deeply branched sediment archaeal phylum with pathways for
acetogenesis and sulfur reduction. ISMEJ 10: 1696 EP –.
Shanks, W.C. (2001) Stable Isotopes in Seafloor Hydrothermal Systems: Vent fluids,
hydrothermal deposits, hydrothermal alteration, and microbial processes. Reviews in
Mineralogy and Geochemistry 43: 469–525.
Shoseyov, O., Shani, Z., and Levy, I. (2006) Carbohydrate Binding Modules: Biochemical
Properties and Novel Applications. Microbiology and Molecular Biology Reviews 70: 283–
295.
Shulse, C.N., Maillot, B., Smith, C.R., and Church, M.J. (2017) Polymetallic nodules, sediments,
and deep waters in the equatorial North Pacific exhibit highly diverse and distinct bacterial,
archaeal, and microeukaryotic communities. MicrobiologyOpen 6: e00428–n/a.
Sim, M.S., Bosak, T., and Ono, S. (2011) Large Sulfur Isotope Fractionation Does Not Require
Disproportionation. Science 333: 74–77.
Spang, A., Saw, J.H., Jørgensen, S.L., Zaremba-Niedzwiedzka, K., Martijn, J., Lind, A.E., et al.
(2015) Complex archaea that bridge the gap between prokaryotes and eukaryotes. Nature
521: 173–179.
Starkenburg, S.R., Chain, P.S.G., Sayavedra-Soto, L.A., Hauser, L., Land, M.L., Larimer, F.W.,
et al. (2006) Genome Sequence of the Chemolithoautotrophic Nitrite-Oxidizing Bacterium
Nitrobacter winogradskyi Nb-255. AEM 72: 2050–2063.
Starnawski, P., Bataillon, T., Ettema, T.J.G., Jochum, L.M., Schreiber, L., Chen, X., et al. (2017)
Microbial community assembly and evolution in subseafloor sediment. PNAS 114: 2940–
2945.
Steen, A.D., Kevorkian, R.T., Bird, J.T., Dombrowski, N., Baker, B.J., Hagen, S.M., et al. (2016)
Extracellular peptidases in subsurface sediments of the White Oak River estuary, NC,
suggest microbial community adaptation to oxidize degraded organic matter.
Stein, C.A. and Stein, S. (1994) Constraints on hydrothermal heat flux through the oceanic
lithosphere from global heat flow. J. Geophys. Res. 99: 3081–3095.
Stein, R. (1990) Organic carbon content/sedimentation rate relationship and its
paleoenvironmental significance for marine sediments. Geo-Marine Letters 10: 37–44.
Sternbeck, J., Sohlenius, G., and Hallberg, R.O. (2000) Sedimentary Trace Elements as Proxies
to Depositional Changes Induced by a Holocene Fresh-Brackish Water Transition. Aquatic
Geochemistry 6: 325–345.
Stokke, R., Roalkvam, I., Lanzén, A., Haflidason, H., and Steen, I.H. (2012) Integrated
metagenomic and metaproteomic analyses of an ANME-1-dominated community in marine
cold seep sediments. Environ Microbiol 14: 1333–1346.
Swan, B.K., Chaffin, M.D., Martinez-Garcia, M., Morrison, H.G., Field, E.K., Poulton, N.J., et
al. (2014) Genomic and Metabolic Diversity of Marine Group I Thaumarchaeota in the
Mesopelagic of Two Subtropical Gyres. PLOS ONE 9: e95380.
Takai, K. and Horikoshi, K. (2000) Rapid Detection and Quantification of Members of the
Archaeal Community by Quantitative PCR Using Fluorogenic Probes. AEM 66: 5066–5072.
Tanaka, T., Fukui, T., Fujiwara, S., Atomi, H., and Imanaka, T. (2004) Concerted Action of
Diacetylchitobiose Deacetylase and Exo-β-D-glucosaminidase in a Novel Chitinolytic
173
Pathway in the Hyperthermophilic Archaeon Thermococcus kodakaraensis KOD1. Journal
of Biological Chemistry 279: 30021–30027.
Tao, S., Eglinton, T.I., Montluçon, D.B., McIntyre, C., and Zhao, M. (2016) Diverse origins and
pre-depositional histories of organic matter in contemporary Chinese marginal sea
sediments. Geochimica et Cosmochimica Acta 191: 70–88.
Teplyuk, A.V., Samarov, N.I., Korzhenkov, A.A., Ulyanova, M.O., Goeva, M.A., Kublanov,
I.V., et al. (2017) Analysis of chitinase diversity in the Baltic Sea bottom sediments.
Microbiology 86: 150–154.
Teske, A., Brinkhoff, T., Muyzer, G., Moser, D.P., Rethmeier, J., and Jannasch, H.W. (2000)
Diversity of Thiosulfate-Oxidizing Bacteria from Marine Sediments and Hydrothermal
Vents. AEM 66: 3125–3133.
Teske, A., Hinrichs, K.-U., Edgcomb, V., de Vera Gomez, A., Kysela, D., Sylva, S.P., et al.
(2002) Microbial Diversity of Hydrothermal Sediments in the Guaymas Basin: Evidence for
Anaerobic Methanotrophic Communities. AEM 68: 1994–2007.
Thamdrup, B., Finster, K., Fossing, H., Hansen, J.W., and Jorgensen, B.B. (1994) Thiosulfate
and sulfite distributions in porewater of marine sediments related to manganese, iron, and
sulfur geochemistry. Geochimica et Cosmochimica Acta 58: 67–73.
Thureborn, P., Franzetti, A., Lundin, D., and Sjöling, S. (2016) Reconstructing ecosystem
functions of the active microbial community of the Baltic Sea oxygen depleted sediments.
PeerJ 4: e1593.
Thureborn, P., Lundin, D., Plathan, J., Poole, A.M., Sjöberg, B.-M., and Sjöling, S. (2013) A
Metagenomics Transect into the Deepest Point of the Baltic Sea Reveals Clear Stratification
of Microbial Functional Capacities. PLOS ONE 8: e74983.
Torti, A., Lever, M.A., and Jorgensen, B.B. (2015) Origin, dynamics, and implications of
extracellular DNA pools in marine sediments. Marine Genomics 24, Part 3 IS -: 185–196.
Trembath-Reichert, E., Morono, Y., Ijiri, A., Hoshino, T., Dawson, K.S., Inagaki, F., and
Orphan, V.J. (2017) Methyl-compound use and slow growth characterize microbial life in 2-
km-deep subseafloor coal and shale beds. PNAS 114: E9206–E9215.
Tully, B.J. and Heidelberg, J.F. (2016) Potential Mechanisms for Microbial Energy Acquisition
in Oxic Deep-Sea Sediments. AEM 82: 4232–4243.
Urich, T., Lanzén, A., Stokke, R., Pedersen, R.B., Bayer, C., Thorseth, I.H., et al. (2013)
Microbial community structure and functioning in marine sediments associated with diffuse
hydrothermal venting assessed by integrated meta-omics. Environ Microbiol 16: 2699–2710.
Valentine, D.L. (2011) Emerging Topics in Marine Methane Biogeochemistry. Annu. Rev. Mar.
Sci. 3: 147–171.
van de Vossenberg, J., Woebken, D., Maalcke, W.J., Wessels, H.J.C.T., Dutilh, B.E., Kartal, B.,
et al. (2013) The metagenome of the marine anammox bacterium “Candidatus Scalindua
profunda” illustrates the versatility of this globally important nitrogen cycle bacterium.
Environ Microbiol 15: 1275–1289.
Vineis, J.H., Ringus, D.L., Morrison, H.G., Delmont, T.O., Dalal, S., Raffals, L.H., et al. (2016)
Patient-Specific Bacteroides Genome Variants in Pouchitis. mBio 7:.
Wakeham, S.G., Lee, C., Hedges, J.I., Hernes, P.J., and Peterson, M.J. (1997) Molecular
indicators of diagenetic status in marine organic matter. Geochimica et Cosmochimica Acta
61: 5363–5369.
Walker, C.B., la Torre, de, J.R., Klotz, M.G., Urakawa, H., Pinel, N., Arp, D.J., et al. (2010)
Nitrosopumilus maritimus genome reveals unique mechanisms for nitrification and
174
autotrophy in globally distributed marine crenarchaea. PNAS 107: 8818–8823.
Walsh, E.A., Kirkpatrick, J.B., Rutherford, S.D., Smith, D.C., Sogin, M., and D’Hondt, S. (2016)
Bacterial diversity and community composition from seasurface to subseafloor. ISME J 10:
979–989.
Wang, F.-P., Zhang, Y., Chen, Y., He, Y., Qi, J., Hinrichs, K.-U., et al. (2013) Methanotrophic
archaea possessing diverging methane-oxidizing and electron-transporting pathways. ISME J
8: 1069–1078.
Wang, H.-L., Zhang, J., Sun, Q.-L., Lian, C., and Sun, L. (2017) A comparative study revealed
first insights into the diversity and metabolisms of the microbial communities in the
sediments of Pacmanus and Desmos hydrothermal fields. PLOS ONE 12: e0181048.
Wang, Y. and Qian, P.-Y. (2009) Conservative Fragments in Bacterial 16S rRNA Genes and
Primer Design for 16S Ribosomal DNA Amplicons in Metagenomic Studies. PLOS ONE 4:
e7401.
Wankel, S.D., Buchwald, C., Ziebis, W., Wenk, C.B., and Lehmann, M.F. (2015) Nitrogen
cycling in the deep sedimentary biosphere: nitrate isotopes in porewaters underlying the
oligotrophic North Atlantic. BG 12: 7483–7502.
Wasmund, K., Mußmann, M., and Loy, A. (2017) The life sulfuric: microbial ecology of sulfur
cycling in marine sediments. Environmental Microbiology Reports 9: 323–344.
Wellsbury, P. and Parkes, R.J. (1995) Acetate bioavailability and turnover in an estuarine
sediment. FEMS Microbiology Ecology 17: 85–94.
Wessel, P., Sandwell, D., and Kim, S.-S. (2010) The Global Seamount Census. Oceanog. 23:
24–33.
Wheat, C.G. and Fisher, A.T. (2008) Massive, low-temperature hydrothermal flow from a
basaltic outcrop on 23 Ma seafloor of the Cocos Plate: Chemical constraints and
implications. Geochem. Geophys. Geosyst. 9: n/a–n/a.
Wheat, C.G., Fisher, A.T., McManus, J., Hulme, S.M., and Orcutt, B.N. (2017) Cool seafloor
hydrothermal springs reveal global geochemical fluxes. Earth and Planetary Science Letters
476: 179–188.
Wheat, C.G., McManus, J., Mottl, M.J., and Giambalvo, E. (2003) Oceanic phosphorus
imbalance: Magnitude of the mid-ocean ridge flank hydrothermal sink. Geophys. Res. Lett.
30: n/a–n/a.
Whitman, W.B., Coleman, D.C., and Wiebe, W.J. (1998) Prokaryotes: The unseen majority.
PNAS 95: 6578–6583.
Wickham, H. (2011) ggplot2: Elegant Graphics for Data Analysis. Biometrics 67: 678–679.
Winter, J.M., Behnken, S., and Hertweck, C. (2011) Genomics-inspired discovery of natural
products. Omics 15: 22–31.
Wu, L., Kellogg, L., Devol, A.H., Tiedje, J.M., and Zhou, J. (2008) Microarray-Based
Characterization of Microbial Community Functional Structure and Heterogeneity in Marine
Sediments from the Gulf of Mexico. AEM 74: 4516–4529.
Yilmaz, P., Parfrey, L.W., Yarza, P., Gerken, J., Pruesse, E., Quast, C., et al. (2013) The SILVA
and “All-species Living Tree Project (LTP)” taxonomic frameworks. Nucleic Acids
Research 42: D643–D648.
Yu, N.Y., Wagner, J.R., Laird, M.R., Melli, G., Rey, S., Lo, R., et al. (2010) PSORTb 3.0:
improved protein subcellular localization prediction with refined localization subcategories
and predictive capabilities for all prokaryotes. Bioinformatics 26: 1608–1615.
Zhang, Ying, Zagnitko, O., Rodionova, I., Osterman, A., and Godzik, A. (2011) The FGGY
175
Carbohydrate Kinase Family: Insights into the Evolution of Functional Specificities. PLoS
Comput Biol 7: e1002318.
Zhang, Yu, Wang, X., Zhen, Y., Mi, T., He, H., and Yu, Z. (2017) Microbial Diversity and
Community Structure of Sulfate-Reducing and Sulfur-Oxidizing Bacteria in Sediment Cores
from the East China Sea. Front. Microbiol. 8: 579.
Ziebis, W., McManus, J., Ferdelman, T., Schmidt-Schierhorn, F., Bach, W., Muratli, J.,
Edwards, K.J., and Villinger, H. (2012a) Interstitial fluid chemistry of sediments underlying
the North Atlantic gyre and the influence of subsurface fluid flow. Earth and Planetary
Science Letters 323-324: 79–91.
Zinke, L.A., Mullis, M.M., Bird, J.T., Marshall, I.P.G., Jorgensen, B.B., Lloyd, K.G., et al.
(2017) Thriving or surviving? Evaluating active microbial guilds in Baltic Sea sediment.
Environmental Microbiology Reports 9: 528–536.

Abstract (if available)

Abstract Sediments lay at the bottom of the ocean, containing the largest active reservoir of organic carbon on this planet (Arndt et al., 2013). Microorganisms survive in these dark environments, recycling and reworking organic matter over geologic timescales (Kallmeyer et al., 2012

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Unexplored microbial communities in marine sediment porewater

PDF

The impact of the concentration and distribution of dissolved and particulate B-vitamins and their congeners on marine microbial ecology

PDF

Microbial ecology in the deep terrestrial biosphere: a geochemical, metagenomic and culture-based approach

PDF

Molecular ecology of marine cyanobacteria: microbial assemblages as units of natural selection

PDF

Patterns of molecular microbial activity across time and biomes

PDF

Thermal diversity within marine phytoplankton communities

PDF

Spatial and temporal dynamics of marine microbial communities and their diazotrophs in the Southern California Bight

PDF

Changes in the community composition of marine microbial eukaryotes across multiple temporal scales of measurement

PDF

Ecophysiology of important understudied bacterioplankton through an integrated research and education approach

PDF

Examining potential triggers of algal blooms and harmful algae in the Southern California bight region

PDF

Application of evolutionary theory and methods to aquatic ecotoxicology

PDF

Disentangling the ecology of bacterial communities in cnidarian holobionts

PDF

The development of novel measures of landscape diversity in assessing the biotic integrity of lotic communities

PDF

The dynamic regulation of DMSP production in marine phytoplankton

PDF

Sex differences in aging and the effects of mitochondria

PDF

Microbial metabolism in deep subsurface sediments of Guaymas Basin (Gulf of California): methanogenesis, methylotrophy, and asgardarchaeota

PDF

Ecological implications of daily-to-weekly dynamics of marine bacteria, archaea, viruses, and phytoplankton

PDF

Annual pattern and response of the bacterial and microbial eukaryotic communities in an aquatic ecosystem restructured by disturbance

PDF

Spatial and temporal investigations of protistan grazing impact on microbial communities in marine ecosystems

PDF

Temporal variability of marine archaea across the water column at SPOT

Asset Metadata

Creator Zinke, Laura Alice (author)

Core Title Microbial communities in marine sediments affecting and effecting biogeochemical cycling: influence of microbial ecology on geochemical transformations in two contrasting marine settings

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Biology (Marine Biology and Biological Oceanography)

Publication Date 08/02/2018

Defense Date 05/09/2018

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag Biogeochemistry,Ecology,marine,microbiology,OAI-PMH Harvest,sediment

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Amend, Jan P. (committee chair), Corsetti, Frank (committee member), Heidelberg, John (committee member), Reese, Brandi Kiel (committee member)

Creator Email lazinke@ucdavis.edu,zinke@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-46264

Unique identifier UC11671416

Identifier etd-ZinkeLaura-6603.pdf (filename),usctheses-c89-46264 (legacy record id)

Legacy Identifier etd-ZinkeLaura-6603.pdf

Dmrecord 46264

Document Type Dissertation

Format application/pdf (imt)

Rights Zinke, Laura Alice

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA