During the spring of 2009, as part of the Ocean–Atmosphere–Sea
Ice–Snowpack (OASIS) campaign in Barrow, Alaska, USA, we examined the
identity, population diversity, freezing nucleation ability of the microbial
communities of five different snow types and frost flowers. In addition to
the culturing and gene-sequence-based identification approach, we utilized a
state-of-the-art genomic next-generation sequencing (NGS) technique to
examine the diversity of bacterial communities in Arctic samples. Known phyla
or candidate divisions were detected (11–18) with the majority of sequences
(12.3–83.1 %) belonging to one of the five major phyla: Proteobacteria,
Actinobacteria, Bacteroidetes, Firmicutes, and Cyanobacteria. The number of
genera detected ranged from, 101–245. The highest number of cultivable
bacteria was observed in frost flowers (FFs) and accumulated snow (AS) with
325
The snowpack has been shown to act as an important matrix for (photo) chemical and biological reactions of organic compounds (Ariya et al., 2011). Snow and ice provide large surface areas which consist of interstitial air, water and ice that may exchange chemical and biological matter with the atmospheric boundary layer. Trace gas exchange, scavenging, photolysis, adsorption (Kos et al., 2014), and more recently, biological transformations in snowpack have been considered (Amoroso et al., 2010, 2009; Fujii et al., 2010; Segawa et al., 2005). Yet, the role of biomolecules, including microorganisms, in oxidation, ice nucleation, gas-particle transfer and aerosol formation remains poorly understood.
Climate change has been linked to changes in snow and ice patterns in the Arctic, potentially impacting the Earth's albedo and atmospheric energy balance (Grenfell and Maykut, 1977; Grenfell and Perovich, 1984, 2004; Hanesiak, 2001). Atmospheric transport events such as dust storms initiated long distances away have been considered to influence the Arctic climate. Saharan dust, for instance, has been reported as a source of certain biological particles to reach the Arctic region (Barkan and Alpert, 2010). In 1976, an Asian dust storm was responsible for bringing as much as 4000 t of dust per hour to the Arctic (Rahn et al., 1977). Dust has also been shown to transport microorganisms (S. Zhang et al., 2007, 2008; X. F. Zhang, 2008). Bacteria and fungi have been detected in Asian dust (Choi et al., 1977; Yeo and Kim, 2002; Wu et al., 2004; Ho et al., 2005) and in African desert winds (Griffin et al., 2001, 2003, 2007, 2006; Kellogg et al., 2004; Prospero et al., 2005), whereby some have been found to be viable (Griffin et al., 2001; Prospero et al., 2005). Recently, the increase in the number of storms has been associated with the efficient long-range transport of dust, microbial and other chemicals to the Arctic regions (Clarke et al., 2001; Grousset et al., 2003; Uno et al., 2009). During long distance transportation, air masses may undergo chemical and physical transformation under extreme environmental conditions such as high levels of solar radiation, multiple freeze–thaw cycles, relatively acidic conditions, and predominantly inorganic salts (Jickells, 1999; Ariya et al., 2002, 2009; Cote et al., 2008). Little is known on the effects of the photochemical and aging processes of the chemical and biological composition of dust particles, or whether chemical properties and the genomic structure of microbial entities transported with dust are altered, or mutated during long distance transport (Smith et al., 2010).
Pure water droplets homogeneously freeze in the atmosphere at approximately
Both natural (e.g. mineral dust, biogenic nucleators) and anthropogenic (e.g. soot) sources can contribute to precipitation in Arctic regions (Hansson et al., 1993; Hinkley, 1994). There is some evidence for the observed increase in the number of storms in certain areas of the globe which can alter the transport and distribution of chemicals or biological entities (Wang et al., 2011; Zhang et al., 2007; Erel et al., 2006) with potential impacts on precipitation patterns (Sempere and Kawamura, 1994; Satsumabayashi et al., 2001). Although the pivotal role of dust in the atmospheric global circulation (Dunion and Velden, 2004; Wu, 2007), radiative budget (Sokolik and Toon, 1996; Kaufman et al., 2001), air pollution (Prospero, 1999; VanCuren, 2003) and cloud formation (Toon, 2003) has been documented, there is little known about how the newly introduced pool of transported microbial entities by dust to the Arctic impacts the change of the total Arctic microbial pool or affects the freezing and melting processes of snow and ice matrices in this region.
Several studies using standard microbiology techniques have shown that there is a diverse population of bacteria in the snow (Carpenter et al., 2000; Amato et al., 2007; Mortazavi et al., 2008; Amoroso et al., 2010; Moller et al., 2011; Liu et al., 2011; Harding et al., 2011). Recent developments in high-throughput sequencing (HTS) techniques (Loman et al., 2012a, b), such as next-generation sequencing (NGS), also allow for metagenomic investigations of microbial populations in environmental samples. The present study was performed as part of the international Ocean–Atmosphere–Sea Ice–Snowpack (OASIS) campaign (2009) in Barrow, Alaska. Five different types of Arctic snow: (i) accumulated snow, (ii) windpack, (iii) blowing snow, (iv) surface hoar snow, (v) fresh snow and frost flowers were used for this study (Fierz et al., 2009; Glossary of Meteorology, 2009). Frost flowers are dendritic shape clusters of ice crystals that form at the interface between warm ice surface and sufficiently cold atmospheric temperature and humidity (Obbard et al., 2009). The chemistry of frost flowers has garnered increased interest because these salty ice crystals have been shown to act as a source for the following: (i) sea-salt aerosol (Perovich and Richter-Menge, 1994), and (ii) BrO, which contributes to ozone depletion events (Kaleschke et al., 2004). Increased bacterial abundance have also been found in frost flowers (Bowman and Deming, 2010). Yet, further research is still required to better understand the mechanisms of physical, chemical and biological processes involving frost flowers.
The aim of the study, in five different snow types and frost flowers in the Arctic, was to evaluate the: (i) identification and quantification of the number of viable bacterial and fungal colonies, (ii) determination of the ice nucleation (IN) property of: (a) selected isolated bacteria and (b) melted samples, and (iii) identification of the total bacterial pool using next-generation sequencing. We herein provide further information on the biological composition of Arctic snow and frost flowers at genomic level, shed light on the potential influence of atmospheric transport on the change of microbial diversity, and discuss their potential roles in the freezing-melting processes of ice-snow in the Arctic.
Five different types of Arctic snow were studied: (i) accumulated snow, (ii)
windpack, (iii) blowing snow, (iv) surface hoar snow, (v) fresh snow and
frost flowers which were collected from 4–20 March 2009 during the
OASIS campaign in Barrow, AK, USA. Detailed snow sampling procedures have
been described elsewhere (Kos et al., 2014). Snow samples
were collected from a field dedicated to snow research in the clean air
sector at 71.31
Snow and frost flower samples were melted directly by transferring from
freezer to refrigerator at 4
Drop-freezing assays were done on: (i) viable isolated bacteria obtained from
Arctic samples, and (ii) melted Arctic samples. Viable bacteria isolated from
Arctic samples grown on Petri dishes were mixed with sterile ultrapure water
(Millipore, Mississauga, Canada). The optical density of 1 at 600 nm was
used to adjust the concentrations of bacteria to 10
Part of each bacterial colony was picked by a sterile disposable inoculating
loop (VWR, Mississauga, Canada) and mixed with
Ready-Lyse™ Lysozyme (Epicentre Technologies,
Madison, USA) and proteinase K in 1.5 m L
The PCR product of 16S rDNA genes obtained from the cultured bacterial colonies were purified using QIAquick PCR Purification Kit (Qiagen, Toronto, Canada), sequenced at McGill University and Génome Québec Innovation Centre, Montreal, Canada. The 16S rDNA sequences were aligned and compared with those available in the GenBank databases using the BLASTN (Basic Local Alignment Search Tool for DNA/nucleic acid) through the NCBI (National Center for Biotechnology Information server) to identify sequences that share regions of homology with isolated sequences.
We opted to use a conventional technique in concentrating the bacteria in
Arctic samples using filtration, sonication and precipitation using
high-speed centrifuge. For DNA analysis, bath sonication is a method that has
been used to dislodge adherent bacteria in environmental samples (Buesing and
Gessner, 2002; Bopp et al.,
2011; Joly et al., 2006; Kesberg and Schleheck,
2013) as well as in medically devised explanted prosthetic instruments
studied in hundreds of patients (Piper et al., 2009; Sampedro et al.,
2010; Tunney et al., 1999). The dislodge bacteria is viable and can be
cultured (Trampuz et al., 2007; Vergidis et al., 2011; Piper et
al., 2009; Sampedro et al., 2010; Tunney et al., 1999; Joly et al., 2006;
Kesberg and Schleheck, 2013; Solon et al., 2011). Melted snow was passed
through a 0.22 micron filter (Millipore, Mississauga, Canada). Filter was
sonicated in 17 m L
The liquid was collected in sterile 50 m L
Analysis transmission electron microscopy (TEM) in conjunction with
energy-dispersive X-ray spectroscopy (EDS) analyses were used on Arctic snow
samples and frost flowers to detect microbial and chemical compounds.
Samples were freeze-dried. A sample solution of 7
Recent observations have indicated that snowpack is indeed a complex microhabitat that permits the growth of diverse microorganisms allowing for photo-chemical and biological reactions to occur (Amoroso et al., 2010). Nitrification (Amoroso et al., 2010), transformation of mercury (Moller et al., 2011) and other pollutants within the snowpack have been detected. Ammonia-oxidizing Betaproteobacteria are active nitrifiers in glacial ice microcosms (Miteva et al., 2007), and the presence of nifH genes has been previously suggested the potential for nitrogen fixation in supraglacial snow (Boyd et al., 2011). A clear understanding of the bacterial population and their interactions will be required to further reveal the role these play in altering the Arctic environment and climate.
In this study, the next-generation sequencing (NGS) technique in conjunction
with a classical cultural method was used to identify and compare the
bacterial community in different types of Arctic snow and frost flowers.
Moreover, the Arctic microbial population was compared to urban snow from
the cold North American city of Montreal. Using GS FLX Titanium (450/Roche),
a total of 88 937 reads was made for all the samples with the average number
of total reads being 17 787. The average read length for all the reading was
a 373 base with an average read quality of 34. After trimming and passing through
quality control, the final read length was recovered as follows: 319
The diversity of the bacterial communities for four different snow types and frost flowers was estimated using the Simpson diversity index, species richness using rarefaction metric, and the nonparametric Chao index (Chao, 1984). The Simpson diversity index which takes into account both species' richness, and an evenness of abundance among the species present reached a plateau after the sequencing of sampling of about 5000 for BS, 6000 for WP, 7000 for US, 8000 for SH, and 10 000 for FFs (Fig. A1a). The Chao index gave values between 1500 and 7500 with BS exhibiting the lowest richness (Fig. A1b). The richness in total bacterial communities of Arctic samples was estimated by rarefaction analysis. The shapes of the rarefaction curves did not reach asymptote, indicating that bacterial richness for most samples especially for urban snow, and windpack is not yet complete (Fig. A1c). Using the 3 % cut-off value in sequence differences for OTU, the estimates of the richness of total bacterial communities ranged from 1033 in BS, 1971 in WP, 1956 in SH, 1933 in US and 1605 in FFs (Fig. A1c). Based on these analyses, the order of the highest diversity of bacteria to the lowest was observed in windpack, surface hoar snow, urban snow, frost flowers, and blowing snow, under experimental conditions herein used.
In the next-generation sequencing part of this study, pyro-sequencing was done only for bacteria and not fungi which was feasible under our existing facilities. However, high resolution electron microscopy (Fig. 1) further revealed the appearance of the existence of several biological materials, remnants of biological activities, and not only biological entities in their entirety. The individual sequences represented known phyla or candidate divisions as follows: 11 (urban snow), Arctic samples: 18 (WP), 16 (SH), 15 (BS), and 18 (FFs) (see also the Appendix Table A1a). The majority of sequences (12.3–83.1 %) belonged to one of the five major phyla: Proteobacteria, Actinobacteria, Bacteroidetes, Firmicutes, and Cyanobacteria. The major phyla for urban snow and Arctic samples were as follows: (i) urban snow – Proteobacteria (49.04 %), Bacteroidetes (47.5 %); (ii) windpack – Proteobacteria (66.1 %), Cyanobacteria (12.3 %); (iii) surface hoar snow – Proteobacteria (67 %), Firmicutes (13.6 %); (iv) blowing snow – Proteobacteria (83.1 %), Firmicutes (6 %), Actinobacteria (5.09 %) and (v) frost flowers – Proteobacteria (50.2 %), and Actinobacteria (32.8 %) (Appendix Table A1a). Proteobacteria was the most widely expressed phylum among all the Arctic samples tested with the greatest abundance observed in blowing snow.
Analysis of snow-associated microorganisms by transmission
electron microscopy (TEM) in conjunction with energy-dispersive X-ray
spectroscopy (EDS). Microorganisms
Bacterial community composition in Arctic samples and urban
snow at genera level as detected by Roche 454 GS-FLX Titanium.
At the genus level, sequences represented 134 different genera for urban
snow; Arctic samples: 245 for windpack, 139 for surface hoar snow, 101 for
blowing snow, and 158 for frost flowers. The distribution of bacterial
genera observed at greater than 1 % and the number of occurrence for each
percentage observed for any genus in total bacteria is shown in Fig. 2.
The name of corresponding genera for each percentage (> 1 %) is
listed in the Appendix Table A2. The top four genera with the highest
percentage detected for each sample were as follows: urban snow (US) –
Bhatia et al. (2006) compared bacterial communities from solid snow and snow melt water from the high Arctic John Evans glacier with basal ice and sub-glacial communities of the same glacier. Distinct bacterial communities were found in each one of these different environments with very few common profiles. Similar to this study, our NGS analysis clearly showed variation of distinct sets of microorganisms among different Arctic samples and urban snow. Our observation also suggests the importance of the selective pressure of specific physical and chemical characteristics of each snow type that may serve as a predictor of microbial abundance and composition (Miteva, 2008). It may specifically favor the growth conditions for microbial communities that originated from diverse sources. Interestingly, few Geobacter bacteria (at 0.09 %) were only detected in the windpack. Some of which have been suggested in previous studies, to catalyze anaerobic U (IV) oxidation with nitrate serving as a potential electron acceptor leading to the subsequent mobilization of uranium (Finneran et al., 2002). Geobacter species have also shown to reduce soluble U(VI) to the less soluble U(IV) (Lovley, 1991). The Arctic region is exposed to further uranium originating from radioactive waste due to military activity, oil and gas, and uranium mining exploitation (Thomas et al., 1992; Dowdall et al., 2004; Convey, 2010; Emmerson and Lahn, 2012). However, additional research is required to evaluate the role of microorganisms in chemical transformation of molecules in the Arctic region.
Relative abundance of origin and physical properties of
analyzed NGS bacteria. Bacteria in Arctic samples: blowing snow (BS), surface hoar snow (SH),
windpack snow (WP), frost flowers (FFs); and urban snow (US) were analyzed
for their origin, and ice nucleation/melting properties as detected by Roche 454
GS-FLX Titanium;
Table 1 shows the analysis of NGS results encompassing bacteria at genus
level that have been previously detected: (i) in Asian or African dust
storms, (ii) with antifreeze and/or ice nucleation properties, and (iii) in
cold oceanic water. Note that the existence of these bacteria does not
ensure the expression of their property, and thus the existence of ice
nucleating and freezing bacteria does not reflect their expression in
environmental matrices. The percentage of bacteria (at genus level) that
were previously observed in Asian or African dust samples were in the range
of 36–47 % for all the snow categories and frost flower samples. Only a
very small percentage of identified bacteria with previously demonstrated
antifreeze property (Yamashita et al., 2002), were
detected in: windpack – 1 %; surface hoar snow – 0.2 %; and frost
flowers – 1 %. Urban Montreal snow samples had the highest number (7 %). 14–16 %
of samples contained bacteria with ice nucleation properties, as shown in
Table 1a. A very small percentage of identified bacteria showed both ice
nucleation and antifreeze properties. Some bacteria such as
Some of the bacteria in Arctic samples have previously been identified in Asian or African storms with ice nucleation (Kellogg et al., 2004; Griffin, 2007) or antifreeze properties (Smith et al., 2013). Within these bacterial genera pool, 2–3 % of Arctic snow samples and urban snow showed ice nucleation properties with only 0.2 % in frost flowers (Table 1b). Bacteria with antifreeze properties were observed for only 0.4 % in frost flowers and 1 % in both blowing snow and surface hoar snow. Higher numbers of such bacteria were observed for windpack (6 %). Interestingly, 13 % of bacteria originating from dust storms in urban snow had antifreeze properties. The possible introduction of antifreeze bacteria from the ocean into the air by different mechanisms such as the bursting of frost flowers by wind and fresh snowfall may further provide and facilitate infiltration into the snowpack (Rankin et al., 2002). The detection of a high number of bacteria with a vast genetic diversity pool, using NGS analysis, further illustrates that the snowpack is a heterogeneous soup of microbial entities. The chemical environment of the snowpack is constantly evolving by novel streams of chemicals through fresh precipitations, wind transportation and metabolic activity of microbial. On a speculative basis, the increased incidence of dust storms, possibly due to climate change, the detection of specific bacteria with possible mid-latitude desert origins into the Arctic environment may suggest a shift in the balance of “native bacterial populations” in the Arctic, yet, there is no current evidence to firmly support this hypothesis and further research is required. One may also speculate that it might be conceivable to consider interactions among the heterogeneous population of microbial in Arctic samples, including non-native taxa adaptation in the Arctic snow-ice genome. Though the Arctic does not provide a native habitat for non-native bacteria or biological species originating from elsewhere in the world, their entrance into the Arctic may affect certain bio-chemical reactions, or alter the nutrient pool for the other native microbial entities. In turn, it might affect the ratio and the survival rate of certain populations of microbial with freezing or anti-freezing properties, impacting the melting of ice or snowpack in the Arctic region. Yet, further studies are required to evaluate such speculations.
With Arctic regions currently warming at rapid rates (Hansen et al.,
2006; Convey et al., 2009), the interrelationship of ice/snow microbial, and
increased water availability is yet to be determined. Though fungi species
and their spore are widespread in the atmosphere, little is known about
their role and presence in the Arctic. Interestingly, a few studies have
shown that fungi like bacteria can be effective ice nucleators, capable of
initiating ice nucleation at temperatures as high as
Only a small fraction of a microbial community, especially from extreme
environments such as the Arctic can be grown under laboratory conditions
since many factors such as the composition of the medium that fully supports
the basic needs of microorganisms for growth is not known. This notion was
further confirmed as cultivable bacteria encompassing 0.1 to 3 % of the
total bacteria, which was detected by NGS technique. Thus, the identified
number of cultivated bacteria and fungi independent from different snow
categories and frost flowers does not reflect the actual number of microbial
and should be considered as the lower limit, and therefore more metagenomic
analysis (such as NGS which was deployed in this study) is essential to
decipher the complex pool of microorganisms in the Arctic. The cultivable
bacteria might be representative of the active fraction of cultivable
bacterial snow communities (Ellis et al., 2003; Frette et
al., 2004), as was detected for bacteria living in different environmental
samples such as soil, and marine samples (Pinhassi et al., 1997; Rehnstam
et al., 1993). Table 2 contains the identified cultivable bacteria found in
each category of Arctic samples (snow and frost flowers). Figure 3a
(bacteria) and 3b (fungi) show the variability in numbers of colony-forming
units (CFUs) within and between the different sample types using two
different media (R2A and TSA) for bacteria and (SDA and mycological) for
fungi. The average number of viable bacteria was higher than the number of
fungi in Arctic samples. Overall, a higher number of CFUs was observed in the
R2A medium with a more limited nutrient content than TSA, wherein most
aerobic bacteria are able to grow. In the R2A plates, the highest number of
bacteria was observed in frost flowers (FFs) and accumulated snow (AS) with
325 and 314 CFU m L
Identification of viable cultivable bacteria in Arctic samples. Some of the bacterial colonies were identified in each snow categories: accumulated snow (AS), blowing snow (BS), fresh snow (FS), windpack (WP), and frost flowers (FFs); accession number, the nearest neighbor found in the database, a unique identifier given to a DNA sequence; identified species; and % similarity, the ratio of identical query bases to known bases in the database.
Concentration of viable cultivable bacteria and
fungi in Arctic samples. Mean number of colony-forming units (CFUs) m L
Both NGS and culture method analysis revealed a very high number of bacteria in frost flowers as compared to the other snow types that we tested. In recent years, special attention has been focused on the role of frost flowers as a contributing factor to changing the chemistry of the atmosphere in the Arctic. Frost flowers are (i) an important source of sea-salt aerosol (Rankin et al., 2002; Perovich and Richter-Menge, 1994; Martin et al., 1995), (ii) a contributing factor in releasing the ozone-depleting molecule, bromine monoxide (BrO), as was detected by satellite (Kaleschke et al., 2004), and (iii) a source of sea ice bacteria (Collins et al., 2010). Moreover, with their physical structure and chemical composition, frost flowers might provide a habitat for microbiological bodies such as bacteria, as well as protective and favorable conditions for metabolic and photochemical reactions (Bowman and Deming, 2010). The observed simple organic compounds and increased concentrations of both formaldehyde (Barret et al., 2009), hydrogen peroxide (Beine and Anastasio, 2009) within frost flower, may suggest that selected bacterial strains can act as a substrate for the photolytic production of oxidants (Bowman and Deming, 2010), and simple organic compounds (Ariya et al., 2002). The regular release mechanism of bacteria through frost flower, such as those with high ice nucleation activity, into the atmosphere, with potential transportation, may provide an additional impact on bioaerosol lower tropospheric mixing ratios (Jayaweera and Flanagan, 1982).
Ice nucleation activity of viable cultivable
bacteria in Arctic samples. The average ice nucleation temperature of:
As opposed to frost flowers, accumulated snow is characterized by several
layers of snowfall, which may have experienced repeated freeze–thaw cycles,
and solar irradiation exposure. Analysis by cultural method showed that the
highest number of bacteria is present in accumulated snow samples. Each fresh
snowfall adds new nutrients and microorganisms to the old pool of accumulated
snow. With the detection of more than 100 organic species in the aerosols at Alert in the
Canadian High arctic (February–June) (Fu et al., 2008), the snow layers
could be further enriched with nutrients by the air/snow exchange (Xie et
al., 2007; Cincinelli et al., 2005). Over time, bacterial populations in
accumulated snow may increase by their sustainability and slow growth at very
low temperature (
Different types of Arctic snow and frost flower samples were tested for IN
activities using obtained cultured bacterial colonies as well as whole
melted Arctic samples. Ultrapure Milli-Q water (18 ohms resistance), tap
water, and
Many of the bacterial isolates in different categories of Arctic samples
showed a moderate IN activity at
Among tested bacterial colonies, fresh snow showed the highest variation in ice nucleation activities; higher variation was observed for accumulated snow as compared to frost flowers. Different factors such as nutrient limitation and low temperature observed in the Arctic might have further shifted the ice nucleation activity of bacteria to the higher temperature (Nemecek-Marshall et al., 1993). Frost flower, a bridge between sea ice and the atmosphere and linking biogenic to non-biogenic materials, had the lowest average ice nucleation activity. This may be related to its salinity and ability to accumulate different chemicals, but further studies are required to provide insight on the physical and chemical processes in frost flowers.
The observed range of IN activity in melted snow and frost flower samples
was between
Ice nucleation of identified viable cultivable bacteria in Arctic sample. Ice nucleation temperature of identified bacterial isolates in each snow categories: blowing snow (BS), fresh snow (FS), windpack (WP), and frost flowers (FFs); accession number, a unique identifier given to a DNA sequence; identified species; and % similarity, the ratio of identical query bases to known bases in the database.
Table 3 enlists the identified cultivable bacteria with their ice nucleation
activity in each category of Arctic samples (snow and frost flowers).
Isolated bacteria were identified belonging to different genus, such as:
The elemental composition of Arctic samples was determined using HR-TEM with EDS. As shown in Fig. 1b, the presence of elements such as Mg, Al, Cl, Ca, and U was detected. The presence of Si and Al might be an indication of soil or through deposit of the dust transported from soil (Sposito, 2008; Shridhar et al., 2010). The source of calcium might be related to either marine or soil origins. Since phosphorous is the limiting macronutrient in marine ecosystems (Toggweiler, 1999; Tyrrell, 1999), the positive observation of this element in frost flower encourages further research in the role of frost flowers in Arctic ecosystem.
It is noteworthy that although we focused on biomolecules materials, as we can see from HR-TEM/EDS analysis, the detection of inorganic matter in Arctic snow and frost flowers can contribute to ice nucleation. Hence, to evaluate the snow freezing properties, the complex chemical and bio-chemical pool of molecules and particles should be considered.
We herein examined the identity, population and ice nucleation ability of the microbial communities of five different snow types and frost flowers during the spring 2009 campaign of the Ocean–Atmosphere–Sea Ice–Snowpack (OASIS) program in Barrow, Alaska, USA. We used the next-generation sequencing (NGS) technique to examine the true bacterial communities in snow and frost flowers, in addition to conventional culture techniques. We gained further insight into the wide range of taxa available in different types of snow and frost flowers. Arctic samples and reference urban snow represented 11–18 known phyla or candidate divisions. The majority of sequences (12.3–83.1 %) belonged to one of the five major phyla: Proteobacteria, Actinobacteria, Bacteroidetes, Firmicutes, and Cyanobacteria. At the genus level, 101–245 different genera were detected. A largely diverse community of bacteria exists in the Arctic with many originating from remote ecological environments such as dust storms. This study revealed that snow and frost flowers are rich media for the existence of microbial compounds. Biological materials have been shown to act as reactive sites for (photochemical) reactions, and thus further studies are required to decipher the complexity of the snow and frost flowers as a zone of chemical pool. It is conceivable that changes on the ratio of antifreeze bacteria to ice nucleation bacteria may have an impact on the melting and freezing processes of snowpack or frost flowers. It is thus feasible that this shift in bacterial population could ultimately affect the snow melting-freezing processes. Further studies are required to evaluate whether change of nucleation patterns due to biological entities are indeed linked to climate change.
Relative abundance of bacterial taxa detected by Roche 454 GS-FLX
Titanium using 16S rDNA gene. Detected phyla
Genus distribution (> 1 %) of Bacterial community in Arctic samples/urban snow by NGS. Windpack snow (WP), surface hoar snow (SH), urban snow (US), blowing snow (BS), and frost flowers (FFs).
Bacterial diversity by 454 NGS analysis. The diversity of the
bacterial communities for Arctic samples: blowing snow (BS), surface hoar
snow (SH), windpack snow (WP), frost flowers (FFs); and urban snow (US) was
estimated using the Inverse Simpson Diversity Index
We thank Gregor Kos for providing us with snow and frost flower samples from Barrow, AK, and Joel Lanoix from Caprion Proteomics Inc. for providing instruments facility. We are also grateful to Paul Shepson of Purdue University, Jan Bottenheim and Sandy Steffen from Environment Canada for logistical support, Florent Dominé and Didier Voisin from the Laboratoire Glaciologie et Géophysique Environnement and Harry Beine from UC Davies for their cooperation during sampling. We thank O. Cavaliere for proofreading the manuscript. Funding from NSERC, CFI, and FRQNT is kindly acknowledged. Edited by: Y. Rudich