UEF//eRepository
DSpace https://erepo.uef.fi
Rinnakkaistallenteet Luonnontieteiden ja metsätieteiden tiedekunta
2019
Archaeal nitrification is a key driver of high nitrous oxide emissions from
arctic peatlands
Siljanen, Henri M P
Elsevier BV
Tieteelliset aikakauslehtiartikkelit
© Elsevier Ltd.
CC BY-NC-ND https://creativecommons.org/licenses/by-nc-nd/4.0/
http://dx.doi.org/10.1016/j.soilbio.2019.107539
https://erepo.uef.fi/handle/123456789/7730
Downloaded from University of Eastern Finland's eRepository
Accepted Manuscript
Archaeal nitrification is a key driver of high nitrous oxide emissions from arctic peatlands
Henri M.P. Siljanen, Ricardo J.E. Alves, G. Ronkainen Jussi, Richard E. Lamprecht, Hem R. Bhattarai, Alexandre Bagnoud, Maija E. Marushchak, Pertti J. Martikainen, Christa Schleper, Christina Biasi
PII: S0038-0717(19)30203-2
DOI: https://doi.org/10.1016/j.soilbio.2019.107539 Article Number: 107539
Reference: SBB 107539
To appear in: Soil Biology and Biochemistry Received Date: 14 March 2019
Revised Date: 1 July 2019 Accepted Date: 10 July 2019
Please cite this article as: Siljanen, H.M.P., Alves, R.J.E., Ronkainen Jussi, G., Lamprecht, R.E., Bhattarai, H.R., Bagnoud, A., Marushchak, M.E., Martikainen, P.J., Schleper, C., Biasi, C., Archaeal nitrification is a key driver of high nitrous oxide emissions from arctic peatlands, Soil Biology and Biochemistry (2019), doi: https://doi.org/10.1016/j.soilbio.2019.107539.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
M AN US CR IP T
AC CE PT ED
ACCEPTED MANUSCRIPT
Archaeal nitrification is a key driver of high nitrous oxide emissions from arctic peatlands 1
2 3
Siljanen Henri M. P.1,2*, Alves Ricardo J. E.2,&, Ronkainen Jussi G.1, Lamprecht Richard E.1, 4
Bhattarai Hem R1, Bagnoud Alexandre2, Marushchak Maija E.1, Martikainen Pertti J.1, Schleper 5
Christa2*, Biasi Christina1 6
7
1University of Eastern Finland, Department of Environmental and Biological Sciences, P.O. Box 8
1627, 70211 Kuopio, Finland 9
2University of Vienna, Department of Ecogenomics and Systems Biology, Archaea Biology and 10
Ecogenomics Division, Althanstrasse 14, A-1090 Vienna, Austria 11
12
&
Present address: Lawrence Berkeley National Laboratory, Climate and Ecosystem Sciences 13
Division, Earth and Environmental Sciences Area, 1 Cyclotron Road, Berkeley, CA 94720, USA 14
15
Keywords: ammonia oxidation; AOA; permafrost; climate change 16
17
* Correspondence to: henri.siljanen@uef.fi, christa.schleper@univie.ac.at 18
19 20 21 22
M AN US CR IP T
AC CE PT ED
ACCEPTED MANUSCRIPT
Abstract 23
Bare peat surfaces created by frost action and wind erosion in permafrost peatlands show high 24
nitrous oxide (N2O) emissions. With global warming, emissions of this highly potent greenhouse 25
gas are expected to increase in Arctic permafrost peatlands. In natural unmanaged soils with low 26
nitrogen deposition, such as Arctic soils, nitrification is the main source of nitrite and nitrate, thus a 27
key driver of N2O emissions. Here, we investigated nitrification, ammonia oxidizer populations and 28
N2O production in vegetated and bare peat soils from four distant Arctic locations. Through a 29
combination of molecular analyses and group-specific inhibitor assays, we show that ammonia 30
oxidation, the first step in nitrification, is mainly performed by ammonia-oxidizing archaea (AOA).
31
All soils from different geographical locations, including bare peat soils with high N2O emissions, 32
harbored only two AOA phylotypes, including an organism closely related to Ca. Nitrosocosmicus 33
spp.. This indicates that high N2O emissions from these ecosystems are primarily fueled by 34
nitrification mediated by very few archaeal species. To our knowledge, Arctic peat soils in this 35
study are the first natural environments where high N2O emissions have been linked to AOA. Any 36
changes in archaeal nitrification induced by global warming will therefore impact on N2O emissions 37
from the permafrost peatlands.
38 39 40 41
M AN US CR IP T
AC CE PT ED
ACCEPTED MANUSCRIPT
Introduction 42
Nitrous oxide (N2O) is the major ozone-depleting compound (Ravishankara et al., 2009) and a 43
greenhouse gas with a global warming potential 298 times that of carbon dioxide (CO2) over a 44
period of 100 years (Myhre et al., 2013). Tropical soils and oceans have long been known to be the 45
major natural sources of N2O (Myhre et al., 2013; van Lent et al., 2015). The role of Arctic 46
ecosystems in the global N2O budget has been largely overlooked, as they are generally N-limited 47
(Christensen et al., 1999; Ludwig et al., 2006; Rodionow et al., 2006; Ma et al., 2007; Takakai et 48
al., 2008) and microbial processes such as nitrification and denitrification are generally assumed to 49
be constrained by low temperatures. However, this view has been challenged by growing evidence 50
that gross nitrification rates in at least some soils are within the same range as those in temperate 51
environments (Booth et al., 2005; Alves et al., 2013), and that some Arctic terrestrial ecosystems 52
may be globally significant sources of N2O – for instance, high N2O emissions have been measured 53
from areas of bare peat on peat plateaus (Repo et al., 2009) and palsa mires (peatlands raised by 54
permafrost) (Marushchak et al., 2011). High N2O emissions from these surfaces are sustained by 55
low carbon (C) to nitrogen (N) ratios (C/N), or high mineral nitrogen availability, the absence of 56
vegetation, and intermediate soil moisture, which allows oxic and anoxic microbial processes to 57
take place simultaneously (Marushchak et al., 2011). Moreover, N2O emissions from both bare and 58
vegetated permafrost peatlands can rise substantially with warming (Voigt et al., 2017a) and 59
permafrost thawing (Voigt et al., 2017b). Together, these observations puts N2O high on the agenda 60
of Arctic research and shows that not only C but also N, should be considered when evaluating 61
climate feedbacks from high-latitude soils.
62 63
Under oxic conditions, N2O is mainly produced in the soil through the ammonia oxidation step of 64
nitrification, whereas under anaerobic conditions, denitrification by heterotrophic microbes is the 65
main source of N2O. Additionally, ammonia oxidizers can also contribute to N2O production via 66
nitrifier-denitrification under low oxygen conditions (Jia and Conrad, 2009; Butterbach-Bahl et al., 67
2013; Zhu et al., 2013; Hu et al., 2015). Laboratory experiments have shown that subarctic bare 68
sub—Arctic peat soils have a high potential for denitrification under anoxic conditions (Palmer et 69
al., 2012), and 15N-site preference values of N2O in these soils have indicated that denitrification is 70
involved in N2O production especially in deeper peat layers (Gil et al., 2017). Consistent with this, 71
denitrifiers were found to be highly abundant in bare peat soils with their community composition 72
being distinct from adjacent vegetated peat surfaces, mainly reflecting the predominance of few 73
specific phylotypes in the former (Palmer et al., 2012). However, denitrification is largely 74
dependent on nitrite (NO2-
) and nitrate (NO3-
) supplied by nitrification, especially in natural 75
M AN US CR IP T
AC CE PT ED
ACCEPTED MANUSCRIPT
unfertilized ecosystems, such as Arctic soils, where N deposition is low and mineralization of 76
organic matter to ammonium (NH4+) is the main source of inorganic N (Reay et al., 2008;
77
Kanakidou et al., 2016). Consequently, nitrification can be considered a key limiting step in N2O 78
emissions from Arctic peatlands, as it regulates the supply of NO2-
and NO3-
for N2O production 79
through denitrification. Moreover, 15N-site preference values of N2O have also identified 80
nitrification as a N2O source, particularly in bare peat during dry years with low surface emissions 81
(Gil et al., 2017). Thus, nitrification has an important dual role in N2O production in bare peat, both 82
indirectly by fueling denitrification and by directly producing N2O (Hu et al., 2015; Kozlowski et 83
al., 2016). Nevertheless, nothing is known about the nitrifying communities and their activity in 84
bare peat soils releasing high amounts of N2O in arctic permafrost ecosystems. Therefore, 85
identification of nitrifier populations and a better knowledge of the factors regulating their activity 86
are paramount to understanding and predicting N2O emissions from these vulnerable ecosystems 87
under current climate change.
88 89
Here, we investigated ammonia oxidizer populations, which catalyze the first and rate-limiting step 90
in nitrification, in permafrost peat soils from four distant Arctic locations spanning Northern 91
Finland, Northwestern Russia, and Northern Central Russia. All bare peat soils at these sites emitted 92
N2O at high net rates, contrary to the adjacent vegetated peat surfaces, which showed very low to 93
negative net N2O emissions (i.e., N2O uptake). Diversity and abundance of ammonia oxidizers were 94
studied based on amoA genes (encoding ammonia monooxygenase subunit A), and the relative 95
contributions of ammonia-oxidizing bacteria (AOB) and ammonia-oxidizing archaea (AOA) to 96
nitrification in a model arctic peatland site were determined using 15N pool dilution assays and 97
group-specific nitrification inhibitors.
98 99
Materials and Methods:
100 101
Study sites and sample collection 102
Soil samples were collected from four Arctic or sub-Arctic peatlands located in northern Russia 103
(Seida, Taymyr and Tazovsky) and northern Finland (Kevo). Bare peat surfaces with high N2O 104
emissions in Seida and Kevo were first identified and characterized by Repo et al. (2009) and 105
Marushchak et al. (2011) (Fig. S1), respectively, whereas the sites in Taymyr and Tazovsky 106
represent previously uncharacterized peatlands. Briefly, bare peat surfaces on permafrost peat 107
plateaus in Seida are approximately 10-500 m2 in area and comprise 3-5% of the total area in the 108
region, where the peat plateau constitutes about 20% of the local landscape (Repo et al., 2009; Biasi 109
M AN US CR IP T
AC CE PT ED
ACCEPTED MANUSCRIPT
et al., 2014). Soil samples were collected during the season of high biological activity (June- 110
August) in 2011 (Seida), 2016 (Kevo), 2011 (Taymyr) and 2012 (Tazovsky) (Table S1.).
111 112
Sampling locations were characterized as bare surfaces and surrounding vegetated surfaces.
113
According to current understanding, bare peat surfaces are created created by wind erosion due to 114
higher elevation compared to surrounding surfaces, or near-surface frost action, or by both of these 115
factors concurrently (Seppälä et al., 2003; Kaverin et al., 2018). At each site, we collected 2-4 116
replicate samples of each surface type (10-30 m apart), and three technical replicates for each 117
(within 0.2-0.5 m distance). Sampling was performed either with a 8 cm x 8 cm box corer, or by 118
using a knife to cut out peat bricks with surface dimensions 15 cm by 15 cm. The top peat layers 119
varied between 5 to 10 cm depth, and were collected depending on peat degradation status. The 120
collected peat samples were stored in re-sealable plastic bags to prevent water loss and kept in 121
cooling boxes with ice bags during transport to the laboratory.
122 123
In the laboratory, soil samples were homogenized by breaking the peat structure by hand. Bags 124
were kept sealed during homogenization, and visible plant material (roots and leaves) and stones 125
were removed using sterilized tweezers. After homogenization, samples were collected in zip-lock 126
plastic bags for further chemical analyses (soil pH, NO3-, and NH4+ contents) using methods 127
previously described (Marushchak et al., 2011; Alves et al., 2013). Sub-samples for molecular 128
analyses were collected in 15 ml plastic tubes and immediately flash frozen with liquid nitrogen.
129
Samples collected at the Russian tundra sites were stored in RNAlater® (Thermo Fisher) during 130
transport to the laboratory. In the laboratory, RNAlater® was removed repeatedly washing soil 131
samples with phosphate-buffered saline buffer solution (1x PBS), using ultra-centrifugation, and 132
washed samples were stored at -80°C.
133
Frozen sub-samples for molecular analysis were ground to a very fine powder using a mortar and 134
pestle with liquid nitrogen. The equipment was sterilized by heating at +400°C for 4 h, or 135
autoclaving, and was also wiped with 70% ethanol. Samples were stored at -80°C until extraction.
136 137
Measurements of nitrous oxide fluxes 138
In situ gas flux measurements were performed using the static chamber method (Marushchak et al., 139
2011; Nykänen et al., 1995) on three different bare peat and adjacent vegetated surfaces within the 140
same peat plateau or palsa mire complex.
141 142
M AN US CR IP T
AC CE PT ED
ACCEPTED MANUSCRIPT
In brief, for in situ gas flux measurements, aluminum collars (dimensions: 60 cm x 60 cm x 15 cm, 143
W x L x H) were installed permanently at the Seida sites in June 2009. Collars were installed on 144
three different bare peat and adjacent vegetated surfaces within the same peat plateau complex. The 145
collar was pressed into the peat down to 5-10 cm depth, and the upper part of the collar had a 146
groove filled with water to create an airtight seal between the chamber and the collar during gas 147
sampling. Each chamber had a circular vent tube to prevent the development of under pressure 148
during sampling (Nykänen et al., 1995). The chambers were equipped with a fan to ensure good 149
mixing of air inside the chamber. Gas samples were collected by closing the chamber (dimensions:
150
60cm x 60cm x 10cm; vol. 36 dm3) for a period of 40 min. At the Kevo and Taymyr study sites, 151
fluxes were measured with a cylindrical stainless-steel chamber (volume ~15 dm3). The open 152
bottom part of the chambers was pushed into the peat to a depth of 3-7 cm, and gas samples were 153
collected at intervals of 5, 10, 20 and 40 min with polypropylene syringes (Terumo®, equipped with 154
three-way stopcocks). Gas samples were transferred into pre-evacuated and N2-flushed glass vials 155
closed with rubber septa within 24 hours (Labco® Exetainer). Air and chamber temperature were 156
measured at the start and end of each measurement.
157
Gas samples were analyzed with an Agilent gas chromatograph (GC) equipped with a Hayesep Q 158
80/100 mesh column (length 1.8 m) and an electron capture detector (ECD) (Nykänen et al., 1995).
159
Flux rates were calculated from the increase in the N2O concentrations with sampling times using a 160
linear regression model. A correlation coefficient > 0.90 was used as quality criterion to accept the 161
flux values.
162 163 164
Soil incubations with the group-specific ammonia oxidation inhibitors.
165
Samples for the inhibitor experiment were collected in triplicate from soil profiles of bare peat in 166
Seida, NW-Russia in August 2014 at depth of 0-10 cm with a box corer. Peat was homogenized by 167
hand mixing, and visible roots and non-degraded plant material were removed. Three microcosms 168
of each replicate (nine in total) were prepared with 20 g of peat in 550 ml incubation bottles.
169
Nitrification in soil was inhibited with allylthiourea (ATU) and carboxy-PTIO (PTIO), which 170
selectively inhibit AOB or AOA, respectively (Shen et al., 2013) The experiment was conducted at 171
natural peat moisture, which was kept constant throughout the incubation period of 30 days. The 172
bottles were closed with rubber septa but the headspace was flushed twice a week with technical air 173
(AGA, Finland) to keep the headspace oxic. Evaporated water due to headspace flushing was 174
compensated. Inhibitor concentrations used in the experiments were selected based on the tests with 175
pure cultures of Nitrososphaera viennensis EN76 (AOA) and Nitrosospira multiformis (AOB) in a 176
M AN US CR IP T
AC CE PT ED
ACCEPTED MANUSCRIPT
previous study (Shen et al. 2013) (note that the actual concentrations of the two specific inhibitors 177
(ATU and PTIO) were different based on the tests). Very high inhibitor concentrations were 178
avoided because at a high concentration these inhibitors lose their specificity against AOB (ATU) 179
and AOA (PTIO). The inhibitors were added twice a week. During the last two weeks of the 180
experiment, the final concentrations of ATU and PTIO were 12 µM and 95 µM, respectively.
181 182
Determination of gross nitrification rates 183
Gross nitrification rates were determined using the 15N pool dilution method (Inselsbacher et al., 184
2007) for each treatment at the end of the inhibitor experiments. Five hundred microliter of 10 at%
185
Na15NO3 (0.301 g l-1) were) was added to 2 g of peat soil in 50 ml flasks, and soils were extracted 186
with 25 ml 2 M potassium chloride (KCl) in duplicates after incubation for 4 h and 24 h (start and 187
end time-points, respectively). Extraction was performed by shaking the peat slurry at 175 rpm min- 188
1 for one hour in plastic 50 ml tubes and filtering the slurry through plastic funnels containing 185 189
mm diameter ashless Whatman® filter paper. Samples for EA-IRMS analysis were prepared with 10 190
ml of extracted sample solution by reducing NO3-
to NH4+
using Devarda´s alloy, after removing 191
ammonium present in the extracts with addition of MgO. This was done by shaking the extracts 192
with acid traps (~6 mm in diameter) at 135 rpm in 20 ml scintillation vials at 35°C for five days.
193
The acid traps were prepared by adding 7.5 µl of 2.5 M KHSO4 to filter paper discs, which were 194
placed and sealed in Teflon tape. After five days of incubation with MgO, the acid traps containing 195
the ammonium initially in the extract were replaced with new acid traps, followed by addition of 196
0.05 g Devarda´s alloy to convert nitrate to ammonium. Samples were shaken at the same speed, 197
temperature and time. Acid traps were gently washed with Milli-Q® H2O and kept in 2 ml 198
microsentrifuge vials with the cap open. The traps were dried for 24 h in a desiccator containing 199
sulphuric acid (≥ 97 %, Sigma-Aldrich). After 24 hours, the traps were opened carefully and folded 200
inside the tin cup for analyses using Isotopic Ratio Mass Spectrometer (Thermo Finnigan), as 201
described in Gil et al., (2017).
202 203
Nucleic acid extraction and purification 204
For DNA extraction, 1 ml (~0.1-0.2 g) homogenized peat soil was transferred into a pre-cooled 205
Lysing tube E (MP Biomedicals, USA) with a sterilized spoon. Lysing matrix tubes were placed in 206
water-ice-bath and 700 µl of preheated (65 °C) CTAB lysing buffer (6 % Cetyltrimethyl 207
Ammonium Bromide (Sigma Life Science, Germany), 1.5 M NaCl 5%, with buffer 250 mM 208
NaPO4, pH 8.0) was added. After addition of buffer-saturated phenol/chloroform/isoamylalcohol 209
(25:24:1; pH 7.5), cell lysis was performed by bead-beating using a FastPrep FP120 device 210
M AN US CR IP T
AC CE PT ED
ACCEPTED MANUSCRIPT
(Thermo Savant, USA) for 30 s at speed 5.5 m s-1. The aqueous phase was extracted with 211
chloroform/isoamylalcohol (24:1) and DNA was precipitated with the 2x volume of 30% PEG6000 212
(Fluka Chemie, Germany) and 1.5 M NaCl solution on ice for 2 h. The tubes were then centrifuged 213
at 4°C for 20 min with 14000 g. The DNA pellet was washed with 1 ml 70% ethanol and then 214
dissolved in 50 µl DEPC-treated H2O (0.01% Diethyl pyrocarbonate, Sigma Life Science, 215
Switzerland). All centrifugation steps were done at 13,200 g for 10 min unless otherwise 216
mentioned. DNA extracts were stored at -80°C until analysis.
217
Extracted DNA was generally not amplifiable by PCR due to high concentrations of co-extracted 218
inhibitory compounds (extracts were visibly yellow to dark brown), such as humic and fulvic acids, 219
which are abundant in peat. Therefore, DNA extracts were column-purified using Sephadex G-50 + 220
15 % PVPP (Fluka/Sigma-Alrich, St Louis USA) and Q-Sepharose (Sigma-Alrich, GE Healthcare, 221
Sweden) (Mettel et al., 2010). Filter columns (0.45 µm pore diameter; GE Lifetechnologies, VWR 222
international) were each filled with 50 µl Sephadex G-50 + 15 % PVPP, followed by sterile DEPC- 223
water, and incubated at room temperature for 1 h. Columns were then centrifuged at 910 g for 5 224
min, washed with 150 µl DEPC H2O and centrifuged again at 910 g for 5 min. A volume of 600 µl 225
Q-Sepharose mixture was added on top of the wet Sephadex/PVPP mixture and columns were 226
centrifuged at 910 g for 5 min and washed twice with 300 µl 1.5 M NaCl.
227
DNA extracts were eluted through the filtration columns in 20 µl volumes with 80 µl 1.5 M NaCl as 228
eluent. Purified DNA was precipitated with 0.1 x volume of 3 M sodium-acetate (10 µl) and 2.5x 229
volume of ice-cold 100 % ethanol (250 µl) at room temperature for 1 h. DNA pellets were washed 230
with 1 ml of ice-cold 70 % ethanol and dried in a vacuum centrifuge at room temperature for 30 231
min. DNA pellets were dissolved in 20 µl DEPC-treated milliQ H2O and stored at -80 °C until 232
analysis. After purification, DNA extracts were clear, without visible dark coloration. DNA 233
concentrations were measured with Qubit® 1.0 Fluorometer (Invitrogen).
234
Samples from the inhibitor incubation for molecular analyses (0.1 g per sample) were collected at 235
the end of the experiment, immediately flash-frozen with liquid nitrogen, homogenized with a 236
mortar and then transferred into Lysing matrix E tubes (MP Biomedicals) tubes. Homogenized 237
samples were stored at -80 °C until nucleic acid extractions. Nucleic acids were extracted and 238
purified with the same protocols described above, but using instead acid phenol/chloroform/IAA 239
(25:24:1, pH 4.5, Ambion 9720) for extraction, and water containing RNase inhibitor (Thermo 240
Scientific RNaseOUT™ Recombinant Ribonuclease Inhibitor 1U µl-1) for preparation of 241
purification columns. Pellets were dried with a SpeedVac for 3 min at 30 °C and resuspended in 50 242
µ L DEPC-treated H2O. Subsamples of the nucleic acid extracts (remaining extracts were kept for 243
DNA-based analyses) were immediately treated with DNAse (1U/ µl, Promega) and incubated with 244
M AN US CR IP T
AC CE PT ED
ACCEPTED MANUSCRIPT
RNase inhibitor (Thermo Scientific) in a heating block at 37 °C for 45 min. Reactions were 245
terminated by adding 1 µl DNAse stop solution and the samples were run for an additional 10 min 246
at 65 °C to stop the enzymatic reaction. Purified RNA extracts were stored at -80 °C until analysis.
247
DNA contamination was checked with archaeal amoA qPCR (see below), and RNA concentrations 248
were measured with the Qubit® RNA HS assay kit (Invitrogen). The yield of DNase-treated RNA 249
was 77 to 112 pg. Complementary (cDNA) was synthesized by incubating random hexamer primers 250
(Thermo Scientific), dNTP mixture and RNA at 65 °C for 5 min, followed by addition of 100 units 251
of RevertAid H Minus Reverse Transcriptase (Thermo Scientific) and 20 units of RNase inhibitor, 252
and incubation at 65°C for 2 h.
253 254
Quantification of amoA genes and transcripts 255
Quantitative PCR (qPCR) of archaeal and bacterial amoA genes was performed using the reaction 256
and cycling conditions indicated in Tables S8 and S9, respectively. All reactions were performed in 257
duplicates. Quantification of archaeal and bacterial amoA genes was based on 10-fold dilutions 258
(101-108) of amoA gene fragments from N. viennensis EN76 and N. multiformis, respectively, 259
amplified from cloned gene fragment with vector specific primers. For archaeal amoA genes, the 260
qPCR efficiency was 76 %, with a detection limit between 6.55x102 and 6.55x107 genes per 261
reaction. For bacterial amoA genes, the qPCR efficiency was 92 %, with a detection limit between 262
1.88x103 and 1.88x108 genes µl-1. The specificity of qPCR amplification products was verified by 263
melting-curve analysis and gel electrophoresis.
264 265
Sequencing and identification of amoA genes 266
Archaeal and bacterial amoA gene fragments of 632 bp and 491 bp, respectively, were amplified 267
with the group-specific primers shown in Table S8 and sequenced with Illumina MiSeq PE250 268
(LGC, Munich, Germany). Given the difficulty in amplifying amoA genes from several samples due 269
to inhibiting humic acids, we tested the PCR amplification efficiency of archaeal amoA genes from 270
each DNA sample using several different DNA polymerases: GoTaq (Promega), DreamTaq 271
(Thermo Fisher), Phusion (Thermo Fisher), Phire II HS (Thermo Fisher). The Phire II HS 272
polymerase, which contains a helping dsDNA-binding domain, yielded the highest amplification 273
efficiency and was thus used in all PCR assays for sequencing. We optimized the PCR conditions 274
for each primer pair by testing different annealing temperatures (56-64.3 °C), primer concentrations 275
(0.156-0.625 µM) and template DNA concentrations (depending on the amount of DNA extracted 276
and soil type). The sufficient amount of DNA template per PCR reaction was selected after testing 277
for potential inhibitory effects on the amplification of archaeal amoA genes. This was done over a 278
M AN US CR IP T
AC CE PT ED
ACCEPTED MANUSCRIPT
dilution series of template DNA spiked and not spiked with a known amount of positive control 279
DNA from all samples to omit the possibility of inhibitory effects of humic acids. The PCR reaction 280
and cycling conditions selected are shown in Tables S9 and S10. Positively amplified (PCR product 281
visible in the agarose gel electrophoresis) of eightfold replicated PCR products were pooled per 282
sample and purified with a PCR product purification kit (Qiagen), according to the manufacturer’s 283
protocol, after verifying the specificity of all PCR products by gel electrophoresis.
284
To check specificity of PCR, unspecific (wrong size) bands were ligated (CloneJET PCR Cloning 285
KIT; Thermo Scientific), cloned into competent cells (One Shot TOP10 kit; Invitrogen) according 286
to manufacturer’s protocol. These clones were sequenced with end-termination chemistry 287
(Macrogen Inc., Netherlands). Obtained correct size AOB PCR products (positive result in 6 out of 288
28 analyzed samples) were sequenced similarly as AOA PCR products with MiSeq PE250, and 289
sequencing reads were quality filtered as described below. The resulted correct and incorrect size 290
bacterial amoA genes were analyzed with BLASTN, (Altschul et al., 1990) and it was shown that 12 291
% of these reads (in total about 140,000 reads were sequenced) were non-specific sequences other 292
than amoA and rest were contaminated with positive control of the laboratory (100 % identical to 293
the amoA gene of N. multiformis).
294 295
Archaeal amoA gene reads were analyzed using the DADA2 workflow (Callahan et al., 2016).
296
Reads were truncated to a length of 200 bp, and those with an expected error greater than 2 were 297
discarded. Given the general lower quality of the reverse reads, only forward reads were used. Non- 298
specific amplicons (i.e., not amoA genes) were identified and discarded by assigning the Amplicon 299
Sequence Variants (ASVs) inferred by DADA2 to the global amoA gene database by Alves et al., 300
(2018) using an identity cut-off lower than 55 % with USEARCH8, (Edgar, 2010) and further 301
inspection of abundant ASVs with BLASTN (Altschul et al., 1990) against the GenBank database.
302
In addition to the read quality and chimera filtering performed by DADA2, additional chimeras 303
were filtered out with UCHIME (Edgar et al., 2011) using the chimera-free reference database by 304
Alves et al. (2018). After quality filtering and chimera removal about 879,000 archaeal amoA gene 305
sequences were classified using the reference database and taxonomy by Alves et al. (2018) with 306
the UCLUST method implemented in QIIME (Caparaso et al., 2010). Data analysis and 307
representation were performed with R 3.4.4. (R Core Team, 2018) using packages Biostrings (Pagés 308
et al., 2017), vegan (Oksanen et al., 2018), string (Wickham, 2018), RColorBrewer (Neuwirth, 309
2014), reshape2 (Wickham, 2007) and ggplot2 (Wickham, 2009).
310 311
M AN US CR IP T
AC CE PT ED
ACCEPTED MANUSCRIPT
We wrote a script that complies the amplicon analysis pipeline used, which can be found at this 312
Github repository: https://github.com/alex-bagnoud/arctic-nitrifiers. The raw primer-free 313
sequencing data were deposited to NCBI SRA under accession number PRJNA488558.
314 315
Statistical analyses 316
All statistical tests were made with R statistical program version 3.4.4 (R Core Team, 2018). Prior 317
to statistical analyses, data were tested for normal distribution using histograms as well as density 318
and qq-plots coupled with the Shapiro-Wilk normality test. To test for correlation between 319
environmental and microbial variables, and N2O fluxes, we applied the Two-Way ANOVA, linear 320
regression model and the non-parametric Spearman’s correlation test. The effect of peat surface 321
type was determined with the Student t-test and pairwise comparisons with Tukey HSD. The effect 322
of inhibitor treatment on gross-nitrification was studied with the linear mixed-effect model using 323
the R package nlme (Pinheiro et al., 2018). We applied non-metric dimensional scaling analysis 324
(NMDS) to reduce data dimensionality and to visualize the variance structure of the dataset, in 325
order to identify differences between sites and surface types in terms of soil archaeal amoA 326
diversity. The effect of sites and surface type on amoA diversity was tested with a PERMANOVA 327
test with adonis function using R package vegan (Oksanen et al., 2018).
328
M AN US CR IP T
AC CE PT ED
ACCEPTED MANUSCRIPT
329 330
Results 331
Soil N2O fluxes 332
We measured N2O emission rates in three out of five Arctic permafrost peatlands located in Kevo 333
(Northern Finland), Seida (Komi Republic, NW-Russia) and Taymyr peninsula (central Siberia, 334
Russia) regions (Fig. 1A, Table S1, Fig. S1), all characterized by significant areas of bare peat 335
surfaces. The three bare peat locations showed high N2O emissions (average flux 28.4 µg N2O m-2 336
h-1, ranging from 5.9 to 42.2 µg N2O m-2 h-1), while emissions from vegetated surfaces were 337
generally low, and occasionally even negative, i.e., N2O uptake (average flux 3.0 µg N2O m-2 h-1, 338
ranging from -1.6 to 6.9 µg N2O m-2 h-1) (Fig. 1B). Nitrous oxide fluxes from bare peat surfaces 339
were lower than those reported earlier from the Seida and Kevo sites (from 2007 to 2013) 340
(Marushchak et al., 2011; Voigt et al., 2017a) (Table 1). However, the fluxes were still high 341
compared with those reported for natural northern soils, and were in the same range as those from 342
tropical and agricultural soils (Martikainen et al., 1993; Christensen et al., 1999; Maljanen et al., 343
2010; van Lent et al., 2015).
344
Soil nitrate content indicated that net nitrification activity (i.e., nitrate accumulation) was higher in 345
the bare peat surfaces (6.64-53.18 µg NO3- g-1 dry soil) than in the vegetated peat surfaces (0.023- 346
0.59 µg NO3- g-1 dry soil) (Fig. 1C), consistent with the gross nitrification rates measured earlier in 347
bare peat soil (Gil, 2017). Soil nitrate content was correlated with the N2O fluxes based on a linear 348
model (F N2O flux vs. NO3- conc.
= 10.91, d.f.1=1, d.f.2=14, P < 0.01, R2 = 0.44).
349 350
Quantification of amoA genes 351
Surprisingly, ammonia-oxidizing archaea (AOA) were the only autotrophic ammonia oxidizers 352
detected across all peat soils from Kevo, Seida, and Taymyr, and two additional sites from 353
Tazovsky (a peat plateau and a peat bog, Western Siberia, Russia). Neither betaproteobacterial 354
AOB or comammox Nitrospira (both clade A and B) were detected based either on end-point or 355
quantitative PCR using group-specific primers for amoA genes. AOA were most abundant in bare 356
surface soils, ranging from 1.81x106 to 1.27x109 amoA genes g-1 dry soil, with the highest 357
abundance at the Taymyr and Seida sites (1.27x109 and 6.39x108 amoA genes g-1 dry soil, 358
respectively) (Fig. 1D). In contrast, archaeal amoA gene abundance was below 4.32x106 genes g-1 359
dry soil in all vegetated soils. On average, abundance of amoA genes was 225 ± 72.5 (mean ± 360
standard error) times higher in bare than in vegetated peat surfaces.
361 362
M AN US CR IP T
AC CE PT ED
ACCEPTED MANUSCRIPT
Sequencing of amoA genes 363
High-throughput sequencing of archaeal amoA genes revealed that AOA communities had very low 364
diversity across all sites, and comprised only two phylotypes of the clades NS-ζ-1.2 and NS- 365
γ−2.3.2 clades, respectively, associated with the taxonomic order Nitrososphaerales (Fig. 2A;
366
amoA-based taxonomy defined by Alves et al. 2018). This was unexpected considering that the 367
samples were collected from very distant arctic geographic locations. Nevertheless, this was 368
consistent with the relatively low diversity of denitrifiers previously observed in bare peat surfaces 369
at the Seida site by Palmer et al., (2012). While clade NS-γ does not have any cultivated 370
representatives, clade NS-ζ corresponds to the recently characterized candidate genus 371
Nitrosocosmicus. The closest cultivated relatives of the most abundant clade NS-ζ amoA gene 372
amplicon sequence variants (ASVs) (Fig. S2) were Candidatus Nitrosocosmicus arcticus Kfb 373
(Alves et al. in review), enriched from a high arctic soil, with 89-92% sequence similarity, followed 374
by Ca. Nitrosocosmicus oleophilus MY3 with 88-90% similarity (Jung et al., 2016). In most soils, 375
the relative abundance of clade NS-ζ was much higher than that of NS-γ, which only dominated at 376
the Tazovsky peat plateau site. The estimated absolute abundance (calculated by multiplying 377
relative abundances of clades by the total amoA gene abundance) of these clades showed that clade 378
NS-ζ dominated AOA communities in Kevo and Seida bare surfaces and in Taymyr vegetated 379
surfaces (Fig. 2B). The fold-difference of clade NS-ζ over NS-γ varied from 2.1 to 20770 in bare 380
soils and from 1.3 to 5.4 in vegetated surfaces (except at the Tazovsky peat plateau in both cases) 381
(Fig. 2C). Soil nitrate concentrations correlated positively with total amoA gene abundance (rho = 382
0.45, P < 0.05) and with the absolute abundance of each clade (rhoNS-γ = 0.45, P < 0.05; and rhoNS-ζ 383
= 0.41; P < 0.05). The abundance of both clades and soil nitrate concentrations had a significant (P 384
< 0.005) interactive effect on N2O fluxes. The absolute abundance of clade NS-ζ together with soil 385
nitrate content had the best explanatory power (F = 14.19, d.f.1=3, d.f.2=12, R2 = 0.78, P < 0.005) 386
for the N2O fluxes measured (Table S6). Despite the low diversity of AOA, archaeal amoA genes 387
exhibited some micro-diversity, as seen among ASVs (Fig. S3), which showed considerable site- 388
related distribution patterns (PERMANOVA (sites) P < 0.05). The consistent micro-diversity 389
between sites also confirmed that the unusual low phylotype diversity across distant geographical 390
locations reflected natural populations and not potential cross-contamination between samples 391
during laboratory analyses (Fig. S2). The average identity between ASVs from clades NS-ζ and 392
NS-γ was 78% (based on consensus cluster sequences).
393 394
Gross nitrification and amoA transcription 395
M AN US CR IP T
AC CE PT ED
ACCEPTED MANUSCRIPT
Gross nitrification rates were approximately 2.5 mg NO3--N kg-1 dry soil h-1 in control soil.
396
incubations (i.e., without inhibitors), which are in the upper range of those measured earlier in 397
Arctic soils (Alves et al., 2013), and about 20-50 times higher than those measured earlier at the 398
Seida site and other tundra ecosystems (Gil, 2017; Wild et al., 2015). Carboxy-PTIO, the AOA- 399
specific inhibitor, significantly reduced gross nitrification rates to approximately 60% of the rates in 400
soils without inhibitor treatments and completely inhibited archaeal amoA gene transcription (Fig.
401
3). In contrast, ATU at concentration known to inhibit only AOB and comammox Nitrospira, had 402
no significant effect on gross nitrification rates, whereas net nitrification was even slightly increased 403
(Fig. S4), consistent with the fact that no AOB or comammox were detected. These results indicated 404
that AOA was specifically inhibited by carboxy-PTIO and that they were the main active ammonia 405
oxidizers in this soil (Fig. 3). Carboxy-PTIO did not, however, lead to full inhibition of nitrification 406
(see discussion below), and did not significantly change the amoA gene abundance, transcription of 407
archaeal 16S rRNA genes, nor soil NO3-
content (Fig. S4).
408 409 410
Discussion 411
In natural soils subject to low inorganic N inputs, such as those in the Arctic, nitrification is the 412
main source of NO2- and NO3-,and is thus also directly or indirectly the first and limiting step of the 413
N2O production. Bare surfaces on permafrost peatlands are known to be significant emitters of N2O 414
in the Arctic (Repo et al., 2009) and recent findings have indicated that global warming enhances 415
these emissions (Voigt et al., 2017a). Here, we identified the key microbial populations driving 416
ammonia oxidation, the rate-limiting step of nitrification, which strongly regulates substrate 417
availability for N2O production across a broad geographical range of arctic peatlands with high N2O 418
emissions. High N2O emissions in these soils are related to optimal soil moisture (0.30-0.70 m3/m-3) 419
and NO3- content (> 5 mg g-1), as previously observed for high N2O-emitting organic soils in the 420
temperate and tropical regions (Pärn et al., 2018).
421 422
Denitrification is generally considered a more important source of N2O than nitrification (Hu et al., 423
2015). Consistent with this assumption, denitrification has been shown to produce high amounts of 424
N2O in bare peat soils from Seida under anoxic conditions (Palmer et al., 2012). Here, we show that 425
archaea were the main active ammonia oxidizers in these soils and thus the main mediators of N2O 426
emissions through nitrite production. Nitrous oxide can then be produced through denitrification of 427
both nitrite and nitrate, and thus archaeal ammonia oxidation not only provides the substrate for 428
M AN US CR IP T
AC CE PT ED
ACCEPTED MANUSCRIPT
nitrate production by nitrite-oxidizing bacteria, but may also feed denitrification directly bypassing 429
the nitrite oxidation step. Moreover, previous 15N-site preference studies have also indicated that 430
nitrification can contribute directly to N2O emissions in these soils, especially during dry seasons 431
(Gil et al., 2017). AOA can produce N2O both under both aerobic and anaerobic conditions through 432
the hybrid formation from NO and hydroxylamine, although production rates are higher under oxic 433
conditions, which prevail in dry soils (Stieglmeier et al., 2013; Kozlowski et al., 2016).
434
Additionally, NO produced by AOA may also generate N2O in peat soils through chemical 435
reactions with abundant humic acids (Zhu-Barker et al., 2015; Kozlowski et al., 2016). Therefore, 436
in addition to fueling denitrification, AOA may also contribute directly to the high N2O emissions 437
from these arctic bare peat soils.
438 439
Ammonia-oxidizer communities in bare and vegetated peat soils were characterized through 440
quantification of amoA genes and gene transcripts, and high-throughput sequencing of amoA genes.
441
While typically both AOA and AOB co-occur in most soils worldwide (Leininger et al., 2006; Jia 442
and Conrad, 2009; Hink et al., 2018), we could only detect archaeal amoA genes across all peat 443
soils studied here, but not those of betaproteobacterial AOB or comammox Nitrospira. Few earlier 444
studies have also shown AOA to be the only detectable or active ammonia oxidizers in some soils 445
(Stopnisek et al., 2010; Herrmann et al., 2012; Isobe et al., 2018), including in soils from high 446
Arctic (Alves et al., 2013) and sub-Arctic (Daebeler et al., 2012) ecosystems. However, to our 447
knowledge, this is the first time that the absence of AOB or sole presence of AOA was observed in 448
soils emitting a high amounts of N2O. Our incubation experiments with group-specific inhibitors 449
confirmed that nitrification activity was indeed carried out mainly by archaea and not by bacteria in 450
peat soils from our long-term experimental site in Seida. Both gross nitrification activity and 451
transcription of amoA genes, were only reduced significantly with the archaea-specific inhibitor 452
carboxy-PTIO, but not with ATU at concentrations known to inhibit specifically AOB (Shen et al., 453
2013; Sauder et al., 2017) and comammox Nitrospira (van Kessel et al., 2015). Full inhibition of 454
nitrification was, however, not achieved with carboxy-PTIO, likely due to low inhibition efficiency 455
in the complex peat matrix. However, we cannot rule out the possibility that heterotrophic nitrifiers 456
might have also contributed to the nitrification activity observed in the presence of carboxy-PTIO, 457
especially by benefitting the inhibition of AOA. While ATU has been shown to affect AOA at 458
concentrations much higher than those used here (Shen et al., 2013; Lehtovirta-Morley et al., 2013), 459
it might have had also a residual inhibitory effect on AOA, which might explain why nitrification 460
was reduced slightly with ATU, although not significantly.
461
M AN US CR IP T
AC CE PT ED
ACCEPTED MANUSCRIPT
462
Surprisingly, we only identified two AOA phylotypes across all geographically-distant arctic 463
peatlands studied here, which effectively represent the whole detectable ammonia oxidizer diversity 464
at these sites. While only few denitrifier phylotypes were also found to dominate denitrifier 465
communities particularly in bare peat surfaces at the same Seida sites, these nevertheless 466
represented a substantially greater diversity of organisms (11-20 OTUs) (Palmer et al., 2012).
467
Likewise, a comparably low diversity of methanotrophic bacteria has also been observed in high 468
arctic wetlands (Graef et al., 2011). These observations indicate that arctic peatlands share common 469
properties that favor the specific AOA phylotypes identified here, but that might have also a more 470
general selective effect on microbial communities. The very low pH of the peat soils studied here 471
(pH ranged from 2.81 to 4.06, Table S1) is likely a major factor selecting for these two specific 472
AOA phylotypes, as previously shown for soil AOA in general (Gubry-Rangin et al., 2011; Alves et 473
al., 2018). Soil pH strongly affects NH3 oxidation by regulating the availability of NH3 (the 474
presumed substrate of ammonia oxidation) in relation to ammonium (NH4+
). Given that NH4+
has a 475
pKa value of 9.25, ammonia availability is extremely limited in highly acidic soils, the 476
predominance of AOA in acidic soils (at least of specific AOA clades) has been proposed to result 477
from their high NH3 affinity (Gubry-Rangin et al., 2011; Prosser and Nicol, 2012). Indeed, the two 478
AOA phylotypes identified here belong to two subclades (NS-γ 2.3.2 and NS-ζ 1.2.) that have been 479
mainly detected in acidic (pH ≤ 6.5) environments (Alves et al., 2018). Low temperatures have also 480
been suggested to support the prevalence of AOA over AOB in alpine and arctic soils (Alves et al., 481
2013; Daebeler et al., 2012; Nemergut et al., 2008; Siciliano et al., 2009; Lamb et al., 2011;
482
Banerjee et al., 2012; Daebeler et al., 2017) and could be a crucial selective factor for archaea as 483
primary ammonia oxidizers in these ecosystems. A recent study has also observed that Ca. N.
484
arcticus Kfb, an AOA abundant in an arctic peatland (Alves et al., 2013) and a close relative of one 485
of the dominant AOA phylotypes identified here (clade NS-ζ), has higher ammonia oxidation 486
activity at temperatures well above those typical of its native arctic environment (>16 °C) (Alves et 487
al., in review). This suggests that the warmer temperatures currently affecting arctic ecosystems 488
(IPCC, 2018) may directly stimulate soil ammonia oxidation by AOA (at least from this lineage), in 489
addition to expected increases in substrate availability through higher N mineralization rates 490
(Schaeffer et al., 2013). Given current predictions of further temperature increases in the Arctic 491
(IPCC), our results suggest that AOA-mediated nitrification could contribute to a significant 492
positive feedback to global warming by fueling higher N2O emissions from arctic peatlands.
493 494
M AN US CR IP T
AC CE PT ED
ACCEPTED MANUSCRIPT
In conclusion, our results show that AOA have a crucial role in nitrification and high N2O 495
emissions from Arctic permafrost peatlands. High soil N2O emissions are usually associated with 496
manipulated soils, like agricultural soils, where AOB are the main drivers of N2O emissions 497
through nitrification of large inputs of N fertilizer (Hink et al., 2018). To our knowledge, Arctic 498
peat soils in this study are the first natural environments where high N2O emissions have been 499
linked to AOA. Given the large differences in ammonia oxidation and temperature kinetics 500
observed among only a few AOA (e.g., Kits et al., 2017; Alves et al., in review), this study also 501
highlights the need to better understand the physiology of AOA that play key roles in critical 502
biogeochemical processes, particularly in sensitive ecosystems such as permafrost soils. This 503
information will be essential to inform biogeochemical models of the N cycle in these systems and 504
improve our predictions of potential non-carbon feedbacks to climate changes.
505 506 507
Acknowledgments:
508
We thank Simo Jokinen for technical assistance. This work was supported by the Academy of 509
Finland [290315], The Kuopio Naturalists’ Society, Federation of European Microbiologist Society, 510
Saastamoinen Foundation., and project P25369 of the Austrian Science Fund (FWF).
511 512 513
Competing Interests Statement:
514
The authors declare that the research was conducted in the absence of any commercial or financial 515
relationships that could be construed as a potential competing or conflicting interest.
516 517
Author contributions:
518
H.S., R.A., P.M., C.S. and C.B., designed the project. R.A., H.S., C.B., R.L., J.R., H.B. and M.M., 519
collected the soil samples and measured N2O fluxes in the field and participated measurements of 520
soil properties. H.S., R.L., H.B., M.M. and C.B., measured gross-nitrification rates and did the 521
microcosm experiment. H.S., J.R., R.L. and A.B., extracted and purified nucleic acids of the 522
samples, amplified DNA with PCR, and prepared the samples for sequencing. H.S., A.B., R.A. and 523
C.S. did bioinformatics and interpreted the results. H.S., wrote the manuscript and all authors 524
discussed and revised the manuscript.
525 526
M AN US CR IP T
AC CE PT ED
ACCEPTED MANUSCRIPT
527 528 529
References:
530
Altschul, S.F., Gish, W., Miller, W., Myers, E.W., Lipman, D.J., 1990. Basic local alignment search tool.
531
Journal of Molecular Biology 215, 403–410.
532
Alves, R.J.E., Minh, B.Q., Urich, T., von Haeseler, A., Schleper, C., 2018. Unifying the global phylogeny 533
and environmental distribution of ammonia-oxidizing archaea based on amoA genes. Nature 534
Communications 9, 1517. doi: 10.1038/s41467-018-03861-1.
535 536
Alves, R.J.E., Wanek, W., Zappe, A., Richter, A., Svenning, M.M., Schleper, C., Urich, T., 2013.
537
Nitrification rates in Arctic soils are associated with functionally distinct populations of ammonia oxidizing 538
archaea. ISME Journal 7, 1620-1631.
539 540
Alves, R.J.E., Kerou, M., Zappe, A., Bittner, R., Abby, S.S., Schmidt, H.A., Pfeifer, K., Schleper, C. Low 541
temperature induces growth uncoupled from ammonia oxidation in the arctic terrestrial thaumarchaeote Ca.
542
Nitrosocosmicus arcticus. (in review).
543 544
Banerjee, S., Siciliano, S.D., 2012. Factors driving potential ammonia oxidation in Canadian Arctic 545
ecosystems: Does spatial scale matter? Applied and Environmental Microbiology 78, 346-353.
546 547
Biasi, C.J.S., Jokinen, S., Marushchak, M.E., Hämäläinen, K., Trubnikova, T., Oinonen, M., Martikainen, 548
P.J., 2014. Microbial Respiration in Arctic Upland and Peat Soils as a Source of Atmospheric Carbon 549
Dioxide. Ecosystems 17, 112-126.
550 551
Booth, M.S., Stark, J.M., Rastetter, E., 2005. Controls on nitrogen cycling in terrestrial ecosystems: A 552
synthetic analysis of literature data. Ecological monographs 75, 139-157.
553 554
Butterbach-Bahl, K., Baggs, E.M., Dannenmann, M., Kiese, R., Zechmeister-Boltenstern, S., 2013. Nitrous 555
oxide emissions from soils: how well do we understand the processes and their controls? Philosophical 556
Transactions of Royal Society of London. Series B, Biological Sciences 368, 20130122. doi:
557
10.1098/rstb.2013.0122.
558 559 560
Callahan, B.J., McMurde, P.J., Rosen, M.J., Han, A.W., Johnson, A.J., Holmes, S.P., 2016. DADA2: High- 561
resolution sample inference from Illumina amplicon data. Nature Methods 13, 581-583. doi:
562
10.1038/nmeth.3869.
563
Caporaso, J.G., Kuczynski, J., Stombaugh, J., Bittinger, K., Bushman, F.D., Costello, E.K., Fierer, N., 564
Gonzalez Peña, A., Goodrich, J.K., Gordon, J.I., Huttley, G.A., Kelley, S.T., Knights, D., Koenig, J.E., Ley, 565
R.E., Lozupone, C.A., McDonald, D., Muegge, B.D., Pirrung, M., Reeder, J., Sevinsky, J.R., Turnbaugh, 566
P.J., Walters, W.A., Widmann, J., Yatsunenko, T., Zeneveld, J., Knight, R., 2010. QIIME allows analysis of 567
high-throughput community sequencing data. Nature Methods 7, 335–336.
568
Christensen, T.R., Michelsen, A., Jonasson, S., 1999. Exchange of CH4 and N2O in a subarctic heath soil:
569
effects of inorganic N and P and amino acid addition. Soil Biology and Biochemistry 31, 637–641.
570 571
Daebeler, A., Bodelier, P., Hefting, M., Rütting, T., Laanbroek, H., 2017. Soil warming and fertilization 572
altered rates of nitrogen transformation processes and selected for adapted ammonia-oxidizing archaea in 573
sub-arctic grassland soil. Soil Biology and Biochemistry 107, 114-124. 10.1016/j.soilbio.2016.12.013.
574
M AN US CR IP T
AC CE PT ED
ACCEPTED MANUSCRIPT
575
Daebeler, A., Abell, G.C.J., Bodelier, P.L.E., Bodrossy, L., Frampton, D.M.F., Hefting, M.M., Laanbroek, 576
H.J., 2012. Archaeal dominated ammonia-oxidizing communities in Icelandic grassland soils are moderately 577
affected by long-term N fertilization and geothermal heating. Frontiers in Microbiology 3, 352 doi:
578
10.3389/fmicb.2012.00352.
579 580
Edgar, R.C., 2010. Search and clustering orders of magnitude faster than BLAST. BMC Bioinformatics 26, 581
2460-2461.
582
Edgar, R.C., Haas, B.J., Clemente, J.C., Quince, C., Knight, R., 2011. UCHIME improves sensitivity and 583
speed of chimera detection. BMC Bioinformatics 27, 2194-2200 doi: 10.1093/bioinformatics/btr381.
584
Gil, J., Pérez, T., Boering, K., Martikainen, P.J., Biasi, C., 2017. Mechanisms responsible for high N2O 585
emissions from subarctic permafrost peatlands studied via stable isotope techniques. Global Biogeochemical 586
Cycles 31, 172–189, doi:10.1002/ 2015GB005370.
587 588
Gil, J., 2017. Microbial processes responsible for the high N2O emissions from sub-Arctic permafrost 589
peatlands and tropical soils as determined by stable isotopes approaches. Dissertations in Forestry and 590
Natural Sciences, University of Eastern Finland. Grano, Kuopio, Finland. ISBN: 978-952-61-2693-7 (PDF) 591
592
Graef, C., Hestnes, A.G., Svenning, M.M., Frenzel, P., 2011. The active methanotrophic community in a 593
wetland from the high Arctic. Environmental Microbiology Reports 3, 466-472.
594 595
Gubry-Rangin, C., Hai, B., Quince, C., Engel, M., Thomson, B.C., James, P., Schloter, M., Griffiths, R.I., 596
Prosser, J.I., Nicol, G.W., 2011. Niche specialization of terrestrial archaeal ammonia oxidizers. Proceedings 597
of National Academy of Sciences of USA 108, 21206-21211.
598 599
Herrmann, M., Hädrich, A., Küsel, K., 2012. Predominance of thaumarchaeal ammonia oxidizer abundance 600
and transcriptional activity in an acidic fen. Environmental Microbiology 14, 3013–3025 doi:10.1111/j.1462- 601
2920.2012.02882.x.
602 603
Hu, H.W., Chen, D., He, J.Z., 2015. Microbial regulation of terrestrial nitrous oxide formation:
604
understanding the biological pathways for prediction of emission rates. FEMS Microbiology Reviews 2015;
605
39, 729-749. doi: 10.1093/femsre/fuv021.
606 607
Hink, L., Gubry-Rangin, C., Nicol, G.W., Prosser, J.I., 2018. The consequences of niche and physiological 608
differentiation of archaeal and bacterial ammonia oxidisers for nitrous oxide emissions. ISME Journal 12, 609
1084-1093.
610 611
IPCC, 2018: Summary for Policymakers. In: Global warming of 1.5°C. An IPCC Special Report on the 612
impacts of global warming of 1.5°C above pre-industrial levels and related global greenhouse gas emission 613
pathways, in the context of strengthening the global response to the threat of climate change, sustainable 614
development, and efforts to eradicate poverty [V. Masson-Delmotte, P. Zhai, H. O. Pörtner, D. Roberts, J.
615
Skea, P. R. Shukla, A. Pirani, W. Moufouma-Okia, C. Péan, R. Pidcock, S. Connors, J. B. R. Matthews, Y.
616
Chen, X. Zhou, M. I. Gomis, E. Lonnoy, T. Maycock, M. Tignor, T. Waterfield (eds.)]. World 617
Meteorological Organization, Geneva, Switzerland, 32 pp.
618 619
Isobe, K., Ikutani, J., Fang, Y., Yoh, M., Mo, J., Suwa, Y., Yoshida, M., Senoo, K., Otsuka, S., Koba, K., 620
2018. Highly abundant acidophilic ammonia-oxidizing archaea causes high rates of nitrification and nitrate 621
leaching in in nitrogen-saturated forest soils. Soil Biology and Biochemistry 122, 220-227.
622
Inselsbacher, E., Cambui, C.A., Richter, A., Stange, C.F., Mercier, H., Wanek, W., 2007. Microbial activities 623
and foliar uptake of nitrogen in the epiphytic bromeliad Vriesea gigantea. New Phytologist 175, 311–320.
624
M AN US CR IP T
AC CE PT ED
ACCEPTED MANUSCRIPT
Jia, Z., Conrad, R., 2009. Bacteria rather than Archaea dominate microbial ammonia oxidation in an 625
agricultural soil. Environmental Microbiology 11, 1658-1671.
626 627
Jung, M.Y., Kim, J.G., Sinninghe Damste, J.S., Rijpstra, I.C., Madsen, E.L., Kim, S.J., Hong, H., Si, O.J., 628
Kerou, M., Schleper, C., Rhee, S.K., 2016. hydrophobic ammonia-oxidizing archaeon of the 629
Nitrosocosmicus clade isolated from coal tar-contaminated sediment. Environmental Microbiology Reports 630
8, 983-992.
631 632
Kanakidou, M., Myriokefalitakis, S., Daskalakisa, N., Fanourgakis, G., 2016. Past, Present, and Future 633
Atmospheric Nitrogen Deposition. Journal of Atmospheric Sciences 73, 2039-2047.
634
Kaverin, D.A., Pastukhov, A.V., Lapteva, E.M., Biasi, C., Marushchak, M., Martikainen, P., 2016.
635
Morphology and properties of the soils of permafrost peatlands in the southeast of the Bol’shezemel’skaya 636
tundra. Eurasian Soil Science 49: 498–511.
637 638
van Kessel, M.A.H.J., Speth, D.R., Albertsen, M., Nielsen, P.H., Op den Camp, H.J.M., Kartal, B, Jetten, 639
M.S.M., Lücker, S., 2015. Complete nitrification by a single microorganism. Nature 528, 555–559.
640 641
Kits, K.D., Sedlacek, C.J., Lebedeva, E.V., Han, P., Bulaev, A., Pjevac, P., Daebeler, A., Romano, S., 642
Albertsen, M., Stein, L.Y., Daims, H., Wagner, M. 2017. Kinetic analysis of a complete nitrifier reveals an 643
oligotrophic lifestyle. Nature 549, 269-272.
644 645
Kozlowski, J.A., Stieglmeier, M., Schleper, C., Klotz, M.G., Stein, L.Y., 2016. Pathways and key 646
intermediates required for obligate aerobic ammonia-dependent chemolithotrophy in bacteria and 647
Thaumarchaeota. ISME Journal 10, 1836-1845.
648 649
Lamb, E., Han, S., Lanoil, B.D., Henry, G.R., Brummell, M.E., Banerjee, S., 2011. A high Arctic soil 650
ecosystem resists long-term environmental manipulations. Global Change Biology 17, 3187–3194.
651 652
Leininger, S., Urich, T., Schloter, M., Schwark, L., Qi, J., Nicol, G.W., Prosser, J.I., Schuster, S.C., Schleper, 653
C., 2006. Archaea predominate among ammonia-oxidizing prokaryotes in soils. Nature 442, 806-809.
654 655
Lehtovirta-Morley, L.E., Verhamme, D.T., Nicol, G.W., Prosser, J.I., 2013. Effect of nitrification inhibitors 656
on the growth and activity of Nitrosotalea devanaterra in culture and soil. Soil Biology and Biochemistry 657
62, 129-133.
658 659
Ma, W.K., Schautz, A., Fishback, L.A.E., Bedard-Haughn, A., Farrell, R.E., Siciliano, S.D., 2007. Assessing 660
the potential of ammonia oxidizing bacteria to produce nitrous oxide in soils of a high arctic lowland 661
ecosystem on Devon Island, Canada. Soil Biology and Biochemistry 39, 2001–2013.
662 663
Martikainen, P.J., Nykänen, H., Crill, P., Silvola, J., 1993. Effect of a lowered water-table on nitrous-oxide 664
fluxes from northern peatlands. Nature 366, 51-53. doi: 10.1038/366051a0.
665 666
Marushchak, M.E., Pitkämäki, A., Koponen, H., Biasi, C., Seppälä, M., Martikainen, P.J., 2011. Hot spots for 667
nitrous oxide emissions found in different types of permafrost peatlands. Global Change Biology 17, 2601- 668
2614.
669 670
Maljanen, M., Sigurdsson, B.D., Guðmundsson, J., Óskarsson, H., Huttunen, J.T., Martikainen, P.J., 2010.
671
Greenhouse gas balances of managed peatlands in the Nordic countries - present knowledge and gaps.
672
Biogeosciences 7, 2711-2738.
673 674