Pole-to-pole biogeography of surface and deep marinebacterial communitiesJean-François Ghiglionea,b, Pierre E. Galanda,c, Thomas Pommierd, Carlos Pedrós-Alióe, Elizabeth W. Maasf,Kevin Bakkerg,h, Stefan Bertilsoni, David L. Kirchmanj, Connie Lovejoyk, Patricia L. Yagerg, and Alison E. Murrayl,1

aUniversity Pierre et Marie Curie (UPMC) 06, Unité Mixte de Recherche (UMR) 7621, Laboratoire d’Océanographie Microbienne (LOMIC), UMR 8222,Laboratoire d’Ecogéochimie des Environments Benthiques (LECOB), Observatoire Océanologique, F-66650 Banyuls/mer, France; bCNRS, UMR 7621, LOMIC,Observatoire Océanologique, F-66650, Banyuls/mer, France; cCNRS, UMR 8222, LECOB, Observatoire Océanologique, F-66650 Banyuls/Mer, France; dInstitutNational de la Recherche Agronomique (INRA), Unité Sous Contrat (USC) 1364, Ecologie Microbienne de Lyon, Université Claude Bernard Lyon I, UMR 5557,CNRS, F-69622 Villeurbanne, France; eDepartament de Biologia Marina i Oceanografia, Institut de Ciències del Mar (CSIC), E-08003 Barcelona, Spain; fNationalInstitute of Water and Atmospheric Research, Kilbirnie, Wellington, 6241 New Zealand; gSchool of Marine Programs, University of Georgia, Athens, GA 30602;hEcology and Evolutionary Biology Department, University of Michigan, Ann Arbor, MI, 48109; iDepartment of Ecology and Genetics, Limnology, UppsalaUniversity, SE-75239, Uppsala, Sweden; jSchool of Marine Science and Policy, University of Delaware, Lewes, DE 19958; kDépartement de Biologie, Institut deBiologie Intégrative et des Systèmes (IBIS), Université Laval, Québec, QC, Canada G1V 0A6; and lDivision of Earth and Ecosystem Sciences, Desert ResearchInstitute, Reno, NV 89512

Edited by David M. Karl, University of Hawaii, Honolulu, HI, and approved September 6, 2012 (received for review May 14, 2012)

The Antarctic and Arctic regions offer a unique opportunity to testfactors shaping biogeography of marine microbial communitiesbecause these regions are geographically far apart, yet sharesimilar selection pressures. Here, we report a comprehensive com-parison of bacterioplankton diversity between polar oceans, usingstandardized methods for pyrosequencing the V6 region of thesmall subunit ribosomal (SSU) rRNA gene. Bacterial communitiesfrom lower latitude oceans were included, providing a global per-spective. A clear difference between Southern and Arctic Oceansurface communities was evident, with 78% of operational taxo-nomic units (OTUs) unique to the Southern Ocean and 70% uniqueto the Arctic Ocean. Although polar ocean bacterial communitieswere more similar to each other than to lower latitude pelagic com-munities, analyses of depths, seasons, and coastal vs. open waters,the Southern and Arctic Ocean bacterioplankton communities con-sistently clustered separately fromeach other. Coastal surface South-ern and Arctic Ocean communities were more dissimilar from theirrespective open ocean communities. In contrast, deep ocean commu-nities differed less between poles and lower latitude deep watersand displayed different diversity patterns compared with the sur-face. In addition, estimated diversity (Chao1) for surface and deepcommunities did not correlate significantly with latitude or temper-ature. Our results suggest differences in environmental conditions atthe poles and different selectionmechanisms controlling surface anddeep ocean community structure and diversity. Surface bacterio-plankton may be subjected to more short-term, variable conditions,whereas deep communities appear to be structured by longer wa-ter-mass residence time and connectivity through ocean circulation.

bipolar | biodiversity | next-generation sequencing | microbial ecology

Global studies of how microbial communities vary in space andalong environmental gradients have highlighted key questions

about the factors that control the distribution of microbes on earth(1, 2). Polar environments remain poorly studied even though theycould help to identify patterns of bacterial biogeography and clarifythe mechanisms responsible for them. Both of Earth’s polar oceansystems have been essentially isolated for millennia by physicalbarriers that limit water exchange with the other oceans (the ArcticOcean by land masses since >60 Ma and the Southern Ocean by astrong current system since ∼20–40 Ma) (3). However, both oceansare subject to parallel extreme physical forces, such as solar irradi-ance that spans from 24 h of sun in the polar summer to 24 h ofdarkness during the polar winter. At the onset of the polar winter,low temperatures result in sea ice formation, whereas during thepolar spring, ice algal blooms followed by ice melt and phyto-plankton production support higher food webs at both poles (4, 5).Despite similar climate drivers acting on the resident biota,

the two polar oceans are dissimilar in several important aspects.

Most notably are differences in freshwater supply to these sys-tems. Although glacial meltwaters flow into the Southern Oceanand, to a lesser extent, the Arctic, several large river systems withlarge continental drainage basins profoundly influence the hy-drology of the Arctic Ocean. Furthermore, the waters of theSouthern Ocean completely surround the continent of Antarc-tica and are driven by the largest and strongest current system inthe World Ocean, the Antarctic Circumpolar Current (ACC).Conversely, the Arctic Ocean is surrounded by Eurasian andNorth American land masses, with its basin perennially coveredby ice and divided by a midbasin ridge (6).The few direct comparisons of microbial life between the polar

oceans have focused on specific taxa such as the haptophytePhaeocystis (7), the foraminiferan Neogloboquadrina pachyderma(8), and bacteria originally isolated from ice such as Polaribacterirgensii (9) and Shewanella frigidimarina (10). Community-widecomparisons using small subunit (SSU) rRNA gene surveys ofplanktonic Archaea reported sequences that are 99% identicalfrom the two poles (11). However, a latitudinal transect of thePacific Ocean using a community fingerprinting method indicateddifferences in the bacterial and eukaryal communities from thetwo poles as well as from tropical and temperate regions (12).Likewise, sea ice microbial communities at the two poles harborclosely related organisms although differ significantly in the rep-resentation of some groups (9, 13). Finally, in silico comparison ofavailable SSU rRNA gene sequences from marine planktonicbacteria indicate bipolar distributions of some ribotypes (14). Insummary, despite the increasingly widespread application of mo-lecular biological approaches to planktonic communities over thelast 20 y, thorough comparisons of microbial communities at thetwo poles are inconclusive because of a lack of comparabledatasets and poor coverage. This limitation highlights the lack ofa standardized approach and reflects the challenges of samplingover a range of locations and depths in both polar zones.

Author contributions: J.-F.G. and A.E.M. designed research; J.-F.G., P.E.G., C.P.-A., E.W.M.,K.B., D.L.K., C.L., andA.E.M. performed research; J.-F.G., P.E.G., T.P., andA.E.M. analyzed data;and J.-F.G., P.E.G., T.P., C.P.-A., E.W.M., S.B., D.L.K., C.L., P.L.Y., and A.E.M. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

Freely available online through the PNAS open access option.

Data deposition: FASTA and processed DNA sequence data have been deposited in thethe Visualization and Analysis of Microbial Population Structures (VAMPS) database,http://vamps.mbl.edu.1To whom correspondence should be addressed. E-mail: Alison.Murray@dri.edu.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1208160109/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1208160109 PNAS Early Edition | 1 of 6

MICRO

BIOLO

GY

Through a collaborative effort initially motivated by the In-ternational Census of Marine Microbes (15) and opportunitiesprovided by the International Polar Year, we tested the hy-pothesis that bacterioplankton communities are the same at bothpoles. We used a suite of 20 Southern Ocean and 24 ArcticOcean samples from both surface and deep waters. Sampleswere processed and analyzed using an identical approach basedon pyrosequencing of the V6 region of the SSU rRNA gene (16).We specifically compared samples from coastal and open oceansand between winter and summer, to test whether or how envi-ronmental conditions and dispersal shape communities in thepolar oceans. Finally, we analyzed an additional 48 samples fromlower latitudes (Fig. 1) to investigate the polar signal in globalmarine bacterial biogeography.

ResultsSimilarity Among Communities. All Southern Ocean (20), ArcticOcean (24), and lower latitude (48) datasets (Fig. 1 and Table S1)were randomly resampled to ensure the number of sequence tagsfrom each sample were equal to the number of tags from thesample with the fewest sequences (9,107). This resulted in 837,844sequence tags (400,708 from the polar samples) falling into 26,902OTUs (0.03 distance threshold) for all samples, with 11,441 fromthe polar datasets. Hierarchical clustering based on Bray–Curtisdissimilarities revealed that the global pattern of pelagic bacterialdiversity was driven by two main factors: vertical distribution andoceanic subregion; analysis of similarity (ANOSIM) showed bothfactors to be highly significant (UniFrac and P test significance,both P < 0.001) (Fig. 2). Additionally, the influence of environ-mental parameters on global changes in bacterial communitystructure was statistically analyzed by canonical correspondenceanalysis (CCA). However, few variables were available across alldatasets and the best model using temperature, depth, and lati-tude explained only 17.2% of the variance (P < 0.05). We alsoinspected the relationship between latitude and the Chao1 rich-ness estimator and between temperature and Chao1 for the sur-face (defined as 0–40 m) and deep (defined as samples >200 m)datasets (Fig. S1). The trends for both the surface and the deepsamples generally decreased with latitude although were notsignificant (Spearman’s R = −0.19, P = 0.20, n = 44; and Spear-man’s R = −0.17, P = 0.34, n = 33, respectively). The relationshipswith temperature were slightly positive, although not significant[surface dataset (n = 44): R = 0.12, P = 0.43; deep dataset (n =33): R = 0.14, P = 0.45].Within the polar surface group (cluster J; Fig. 2), Southern

and Arctic Ocean bacterioplankton communities always clus-tered apart from each other. Southern and Arctic Ocean summercoastal communities were 65% dissimilar (clusters A and B),winter coastal communities were 55% dissimilar (clusters C and

D), and open ocean communities were 46% dissimilar (clusters Fand G). Surface layer polar communities were highly dissimilar(86%) to those from lower latitudes (clusters J and K; Fig. 2).Bacterial communities in polar deep waters were also different

(72% dissimilar) from those in temperate zones. Water-massconnectivity was also evident; for example, the deep WesternArctic communities (cluster H) clustered closely with bacteria(44% dissimilar) from the North Atlantic Ocean sites, whichwere influenced by Labrador Sea water masses. Furthermore,bacterial communities from the deep Southern Ocean clusterwith communities retrieved from the Eurasian Arctic (cluster I),indicating that there were also fewer differences between polesfor deep samples. However, both communities clustered apartfrom each other (41% dissimilar; Fig. 2).

Phylogenetic Diversity Accumulation in Surface and Deep Communities.To further test factors that could influence the biogeography ofmarine bacterial communities, we calculated phylogenetic diversityaccumulation curves (Fig. 3) in which the number of OTUs vs.the sequence similarity threshold used to define the OTUsare plotted. The curve shape indicates the degree of phylogeneticrelatedness (or distance) among taxa within each community. Iftaxa in a community have low SSU rRNA gene sequence di-vergence, shown by shorter branch lengths in the phylogenetictree, the diversity accumulation curve is expected to indicatea higher percentage of OTUs at high similarity thresholds andfewer OTUs at the lower similarity thresholds than for commu-nities containing more distantly related taxa. Interestingly, differ-ences in the diversity accumulation curves were largest betweensurface and deep communities (ANOSIM R = 0.36; P < 0.0001;n = 92), with significantly fewer OTUs at low SSU rRNA genesequence similarity in the surface compared with deep waters (Fig.3). In the surface samples, a rapid increase in the number of OTUsappeared at the sequence similarity threshold ≥97% identity (Fig.3, break in slopes). This indicates lower divergence in the SSUrDNA gene sequences in surface compared with deep waters. Deepcommunities were more variable than surface communities in theirdiversity accumulation curve shapes, with a greater percentage ofOTUs remaining at low sequence similarity thresholds. The polardeep samples showed the highest degree of diversity accumulation.

Unique and Shared OTUs Between Polar Microbial Communities.When we compared the surface polar oceans to each other (ata 0.03 distance threshold), 78% of the Southern Ocean and 70%of the Arctic Ocean OTUs were exclusive to each pole. For thedeep samples, 45% of the OTUs were unique to the SouthernOcean, whereas 85% were unique to the Arctic Ocean wheremore deep samples were surveyed (Fig. S2). When the entiredataset including lower latitude stations was considered, 70% of

Fig. 1. Polar oriented maps of sample locations inthe Southern (A) and Northern (B) hemispheres,including Southern Ocean (pink), Arctic (blue), andlower latitudes (gray).

2 of 6 | www.pnas.org/cgi/doi/10.1073/pnas.1208160109 Ghiglione et al.

surface Southern Ocean OTUs and 39% of deep SouthernOcean OTUs were never found in the Arctic or the temperateoceans. Similarly for the Arctic Ocean, 61% of surface and 60%of deep were only found in the Arctic (Fig. S3). Furthermore,59% of the shared polar surface and 26% of the shared polardeep OTUs were not present in the lower latitude samples. Only2% of OTUs were found in all surface waters and 4% of OTUswere found in all deep waters.Most OTUs in both polar oceans belonged to Gammapro-

teobacteria, Alphaproteobacteria, and Flavobacteria. However,there were some differences for other groups. For example,Betaproteobacteria, Actinobacteria, and Acidobacteria were morecommon in the Arctic (3–4% each; Fig. S2) than in the SouthernOcean (less than 2% each). Compared with surface, deep watersfrom both polar regions had a lower proportion of Flavobacteria(8% at deep and 14% in surface) and a higher proportion ofDeltaproteobacteria (9% at deep and 3% in surface).

Samples were then grouped at the level of major ecosystemsderived from the clusters in Fig. 2. Using the polar data only, weidentified the most frequent OTUs (28 in total) that accountedfor ≥5% of the sequences in each of the polar ecosystem clusters:Coastal Summer, Coastal Winter, Open Ocean Summer, andDeep Arctic Ocean Surface (Fig. 4). Independently, we identi-fied the OTUs primarily responsible for dissimilarities betweenpairs of these clusters (e.g., Southern Ocean vs. Arctic coastalsummer) as well as between polar and mid latitude samples (Fig.4) using similarity percentage analysis (SIMPER). In most cases,these OTUs were a subset of the 28 most represented OTUs(Fig. 4). Among those top 28 OTUs, 14 were unique to polarareas and not found in the lower latitude samples; see thecomparison of all polar surface with midlatitude clusters J vs. K(Fig. 4). The most important polar bacteria OTUs according toSIMPER that influenced the dissimilarity between ecosystemclusters were SAR11-Pelagibacter-Cluster1, OMG-Ant4D3-Cluster1

SP_2_O110 SP3_O120

SP_5_O145 SP_4_O150 SP_6_O200 NA_8_O100

NA_1_O47 NA_3_O73

AZ_13_O100 SP_7_O110 SP_8_O100

NA_12_O1300 NA_7_O7

NA_10_O11 NA_13_O7 NA_4_O89 NA_9_O48 NA_5_O7 AZ_1_O0 AZ_7_O0 AZ_3_O0 AZ_6_O0 AZ_8_O0

AZ_14_O0 SA_6_O28*

SA_16_O48* MS2_C5 MS6_O5 MS4_O5 MS3_C5 MS5_O5

AO_BS1_C2 AO_CS4_C2

AO_CS16_C2* AO_FB13_C3*

SO_AP2_C6* SO_AP3_C6*

SO_AS10_C250

SO_WS12_C11 SO_RS16_C10

SO_KI6_C3 SO_KI7_C3

SO_KI5_C3* SO_KI8_C3*

SO_WS9_O11 SO_WS10_O11 SO_WS11_O11 SO_RS14_O10 SO_RS15_O10 SO_RS13_O10

AO_CS5_O10 AO_BS7_O10 AO_CS6_O10 AO_BS8_O40

AO_BS3_O1000 AO_BS5_O1000 AO_BS9_O1000

AO_BS12_O1000 AO_BS11_O900 AO_BB1_O1000 AO_BS7_O400 AO_BS4_O400

AO_BS10_O400 AO_BS13_O400

NA137_O1700 NA53_O1400 NA115_O550 NA138_O700 NA55_O500

AO_ES14_O1000 AO_ES15_O1200 AO_ES16_O250 SO_AS13_C790 SO_AS14_C500

AZ_9_O3660 AZ_10_O2100 AZ_12_O800 AZ_11_O100 AZ_15_O800 AZ_16_O100 NA_2_O1100

MS9_O500 MS10_O2000

MS11_C500 SA_14_O4600* NA112_O4000

90 80 60 40 20 0

-Atlantic Ocean Transect (NA) -Azores Water Project (AZ)

-South Pacific Gyre (SP) -Blanes Bay Microbial Observatory (MS)

- Azores Water Project (AZ) - Blanes Bay Microbial Observatory (MS)

-AOT: Atlantic Ocean Transect (SA-NA)

- North Atlantic Deep Waters (NA)

AO_CS2_C2 AO_CS3_C2

AO_FB15_C3

SO_AS1_C10 SO_AP1_C6 SO_AP4_C6

(A) SO coastal summer

(B) AO coastal summer

(C) SO coastal winter

(D) AO coastal winter

(E) Sub-SO coastal

(F) SO open ocean summer

(G) AO open ocean summer

(I) Eastern AO and SO deep

(J) all polar surface

(K) mid-latitude

(H) Western AO deep summer

70 50 30 10

Fig. 2. Unweighted pair groupmethodwitharithmeticmean(UPGMA)dendrogram based on Bray–Curtis dis-similarities of V6 16 rRNA tags from pe-lagic samples taken in Arctic (blue),Southern Ocean (purple), and in lowerlatitudes (gray). The x axis representsBray–Curtis dissimilarity. Ecosystemclusters are representedwith labels andlettersA through I. Samplenamesat leftare encoded as ocean region, VAMPScode number, open (O) or coastal (C)sites, and water depth in meters. Polarocean regions in the Southern Ocean(SO) are Amundsen Sea (AS), AntarcticPeninsula (AP), Kerguelen Islands (KI),Ross Sea (RS), and Weddell Sea (WS). Inthe Arctic Ocean (AO), regions sampledare Baffin Bay (BB), Beaufort Sea (BS),Chukchi Sea (CS), East Siberian Sea (ES),and Franklin Bay (FB). A sample namefollowed by an asterisk refers to wintersampling.

Ghiglione et al. PNAS Early Edition | 3 of 6

MICRO

BIOLO

GY

and -2, and SAR86-Cluster1, and Polaribacter. Further in-spection revealed differences between poles for different envi-ronments (coastal and open ocean) and seasons (summer andwinter). In summer coastal waters, OMG-Ant4D3-Cluster1,Gammaproteobacterium HTCC2207-Cluster 1, and Sulfitobacterwere more abundant in the Southern Ocean, while Polaribacter,Burkholderiales, and Microbacteriaceae OTUs were more abun-dant in the Arctic (Fig. 4). In winter, there were also large dif-ferences between the poles, with OMG-Ant4D3-Cluster1, GSO-EOSA1Complex/Ant10A4, and Methylophaga explaining themajority of the difference for the Southern Ocean, whereasSAR11-Pelagibacter-Cluster1 and OMG-Ant4D3-Cluster2 weremore abundant in the Arctic Ocean. Polar winter coastal com-munities had a higher frequency of SAR324 and uncultivatedGammaproteobacteria OTUs compared with summer samplesfrom the same regions.Polar summer coastal communities were easily distinguished

from summer open ocean communities [Figs. 2 and 4; similarityprofile testing (SIMPROF) analysis]. For example, the coastalcommunities were dominated by the OMG-Ant4D3-Cluster1,along with fewer Polaribacter-, Sulfitobacter-, and Loktanella-af-filiated OTUs, whereas the open ocean samples contained ahigher proportion of SAR11, uncharacterized RoseobacterNAC11-3, and SAR86-Cluster1 and -2. Summer polar surfaceopen ocean samples (Fig. 4, columns F and G) could be distin-guished by the uncharacterized Roseobacter/NAC11-3 OTU,SAR86-Cluster1, uncharacterized Flavobacteriaceae-Cluster3, andan Ulvibacter-related OTU, which were dominant in the SouthernOcean, whereas the SAR11-Pelagibacter-Cluster1 and SAR86-Cluster2 were more abundant in the Arctic. The deep polar oceanclusters (Fig. 4, columns H and I) differed the most because of theSAR324 OTU being more abundant in theWestern Arctic than inthe Eastern Arctic and Southern Ocean. The SAR11-Pelagibacter-Cluster1 and -Cluster2, OMG-Ant4D3-Cluster2, and GSO-EOSAA1Complex/Ant10A4 OTUs also contributed to the dis-similarity between the clusters. The typical deep water cladeSAR406 was equally represented at both poles.

DiscussionThe two polar oceans share similar extreme environmental con-ditions that are very different from other oceans. Moreover, polarocean circulation patterns and cold seawater temperatures havebeen present for up to 25 million years (3). Few studies havecompared bacterioplankton communities at the poles; one tran-sect along the axis in the Pacific Ocean from the Arctic to theSouthern Ocean used a DNA fingerprinting technique, which

indicated differences between the two poles, as well as mid lat-itudes (12). That study detected, but did not identify, ∼11 taxa persample and covered only a small area of each polar zone. Becauseseasonal, regional, and ecological differences also influencebacterial communities, the extent of differences between thepoles required deeper, more extensive, sampling coverage to in-clude the less abundant and rare taxa that make up a substantialportion of bacterial communities in the oceans. Our analysis,based on over 830,000 V6 rRNA gene sequences from 92 samplesdistributed among multiple polar and lower latitudes areas atdifferent times of the year and including surface and deep waters,showed that there were profound differences between Southernand Arctic Ocean bacterioplankton communities.Limited dispersal capacity could account for differences be-

tween polar areas. If bacteria do not remain viable during long-range transport, communities would differentiate from eachother according to a classical allopatric model. However, geneflow between the poles has been suggested for other micro-organisms and benthic foraminifera (17) and very similar(≥99%) SSU rRNA gene sequences belonging to an ice di-noflagellate (18), Planctomycetales (9), other bacteria (14), andmarine group I Crenarchaeota (11) are suggestive of bipolardistributions. We found that about 15% of the OTUs werecommon to both poles, including representatives of the domi-nant Alphaproteobacteria, Gammaproteobacteria, and Flavobac-teria groups that are cosmopolitan in marine waters. SeveralOTUs matched cultivated species from sea ice, including thepsychrophile Polaribacter sp. and the psychrophilic gas-vacuolatebacterium Octadecabacter sp. (19). The potential bipolar distri-bution of a subset of the bacterioplankton is intriguing, and ourresults point toward lineages that warrant further investigation.At the same time, we found that the majority (85%) of the OTUsappeared to have pole-specific distributions, suggesting in-complete dispersal between the poles; geographical isolation,physiological characteristics, and ecological traits could worktogether inhibit widespread bacterioplankton dispersal.Environmental filters may act strongly at the two poles with

greater community differences occurring where selection isgreatest. One difference between the poles is the terrestrial in-fluence on coastal areas. Arctic surface waters are modified bylarge rivers that bring in significant sediment loads and dissolvedorganic carbon, whereas freshwater flows into the SouthernOcean are smaller and mostly from glaciers. The differentqualities and quantities of freshwater drive differences in nutri-ent regimes. For example, we would expect greater differences incoastal than in the open ocean when comparing the two poles.Accordingly, although coastal surface communities of the Arcticand Southern Oceans were more similar to each other comparedwith the open oceans, they were distinct (65% Bray–Curtis dis-similarity in summer) from each other. Indeed, there was greaterrepresentation of some freshwater bacteria including Betapro-teobacteria and Actinobacteria (20) in the surface Arctic Ocean.We found fewer differences between the two polar coastalregions in winter (55% Bray–Curtis dissimilarity), consistent withthe lack of river runoff in the Arctic during winter, and presumedenvironmental convergence of polar winter conditions. Ourresults support the notion that environmental conditions struc-ture bacterial communities of surface polar waters strongly, inparticular in the coastal areas.Although our results are consentient with the emerging global

view of horizontal differences in bacterial communities acrossoceans (e.g., refs. 2 and 21), vertical structure is more strikingand evident in essentially all other oceans including the Pacificand Atlantic Oceans (22, 23) and the Southern Ocean (24). Inthe absence of light and more quiescent conditions, environ-mental drivers may be weaker compared with those in surfacewaters. In support of this, we found that surface communitiesdiffered more than deep communities (Fig. 2): a finding consis-

AO surface AO deep SO surface SO deep

Deep Surface

30

40

50

60

70

80

90

100

100 99 98 97 96 95 94 93 92 91 90 89 88 87

OTU

s (%)

Similarity (%)

MS surface MS deep AZ surface AZ deep NA surface NA deep SP surface

Fig. 3. Diversity accumulation in deep vs. surface communities representedby the percentage of OTUs in a dataset when the sequence similaritythreshold used to define OTUs increases. Datasets included MediterraneanSea (MS), Azores (AZ), North Atlantic (NA), South Pacific (SP), Arctic Ocean(AO) and Southern Ocean (SO).

4 of 6 | www.pnas.org/cgi/doi/10.1073/pnas.1208160109 Ghiglione et al.

tent with environmental rather than geographical isolation beingthe main determinant of the bacterial community composition insurface waters. In addition, when we addressed the question ofwhether there was a relationship between community richnessand latitude, as has been reported by others (e.g., refs. 14 and 25)with lower levels of resolution or depth, we found little supportfor a relationship. Admittedly, the coverage, in even this largestudy, is not adequate to address this intriguing question suffi-ciently. The data analyzed here offer a window into significantvariation within latitude (e.g., 40° and 64°) and between seasons,in particular, at the high latitudes, where winter samples weremore diverse than their summer counterparts.An intriguing aspect of deep water bacterioplankton in this

study was the geographic patterns of similarity among commu-nities. Deep bacterial communities from the East Siberian Seaclustered with deep Southern Ocean communities from theAmundsen Sea but not with other deep Arctic communities(cluster I in Fig. 2). The deep community connection betweenpoles could be explained by the export of deep water formed inthe Eurasian Arctic Basin (Eastern Arctic), which exits the ArcticOcean east of Greenland through the 2,600-m-deep Fram Strait.This dense water moves relatively fast through the Atlantic Oceanalong the deep conveyor belt, reaching the Southern Ocean onthe order of hundreds of years (26). Because environmentalconditions in this water mass would be relatively stable, the deepArctic bacterial communities could be maintained.It is also noteworthy that even with this larger dataset, the

eastern and western Arctic samples did not cluster together (27).These deep basins are separated by the Lomonosov Ridge, whichrises to 650 m below the surface of the ocean (6), thereby sepa-rating the two deep Arctic Ocean Basins. The Canadian ArcticOcean bacterial communities from the deep Beaufort Sea anddeep Baffin Bay clustered together along with North AtlanticDeepWater that is derived from Labrador Current waters (clusterH in Fig. 2; Western Arctic Deep Summer). The deep bacterialcommunities from the Canadian Arctic were from the BeaufortGyre, the cold, dense waters of which eventually exit the ArcticOcean through Baffin Bay. These waters are then exported fromBaffin Bay to the North Atlantic Water Mass sampled here (28).

This global connectivity implies that the biogeography of deep-water communities may be largely controlled by ocean circulation.Deep polar communities were dominated byOTUs in the SAR324cluster ofDeltaproteobacteria, a typical deep ocean phylotype (e.g.,ref. 29). Notably, this OTU and, to a much smaller degree,SAR406 (average 1.8–4.7% compared with 0.1–0.3%) was alsorecovered in surface winter samples from both poles, suggesting atleast some microbes can persist in cold dark waters irrespectiveof depth.Surface and deep communities had different phylogenetic di-

versity patterns, as defined by diversity accumulation curves (30)based on the slowly evolving SSU rRNA gene. The curves pro-vide a general framework for understanding patterns of phylo-genetic diversity. Despite the relatively short length of the V6region of the SSU rRNA gene, we found that levels of sequencedivergence differ substantially between the surface and deeppelagic environments, with less divergent SSU rRNA gene poolsin surface communities. This is consistent with previous reportsof high phylogenetic turnover of surface communities at theglobal scale (31). Such a depth difference argues for differentspeciation patterns and perhaps different evolutionary pressuresshaping community structure in surface vs. the deep waters,because phylogenetic diversity should be lower in a region withrecent and more rapid speciation events and higher in a regionwith slower speciation rates (32). We propose that the differ-ences in phylogenetic diversity between surface and deep polarcommunities are attributable to (i) rapid speciation at the sur-face as consequence of dynamic surface condition, in contrast to(ii) spatially and temporally constant environmental conditionsover longer timescales in the deep sea. These longer scales wouldallow more time to select for phenotypes adapted to the specificconditions, with concomitant loss of less adapted phylotypes.Phylogenetically distant populations would be expected to co-occur if most speciation followed allopatric speciation, whereasco-occurrences of phylogenetically closely related populationswould be controlled by sympatric speciation (33). The more di-vergent OTUs found in deep waters suggest communities shapedby allopatry, whereas the surface communities, with less diverg-

Fig. 4. Bacterial OTUs associated with the polar ecosystemclusters identified in Fig. 2 (clusters A to H). SSU rRNA genetags (distance of 0.03) were averaged across each ecosystemcluster (3–29 samples per cluster) and summed across theeight polar ecosystems. Shown are OTUs representing 5%or greater in each compared ecosystem cluster (totaling 28OTUs), where the circle size corresponds to the relative av-erage abundance the OTU in each cluster. OTUs assigned tothe highest taxonomic level possible using a Bayesian clas-sification tool (RDP) BLAST and sequence alignments topolar SSU rRNA gene clone libraries, are grouped phyloge-netically on the vertical dimension. OTUs with average rel-ative abundances < 0.01 (rare OTUs in a fraction of thesamples in an ecosystem cluster) are indicated with an x,whereas very low values (0.01–0.5) appear as a continuationof dashes. Pairwise comparisons of ecosystem clusters(delimited with vertical black lines and gray background)were conducted using SIMPER to determine OTUs contrib-uting to dissimilarity between the clusters. Dominant tags(top five to six OTUs) that contributed to dissimilarity be-tween two clusters are in white. For example, the six fre-quent OTUs explaining the dissimilarity between the coastalsurface summer samples in the Antarctic and the Arctic in-cluded three Antarctic OTUs: Sulfitobacter, OMGAnt4D3-cluster1, Gammaproteobacterium HTTC 2207-Cluster1, andthree Arctic OTUs: Burkholderiales, Polaribacter, andMicrobacteriaceae); the discerning OTUs are often distrib-uted between the two clusters being compared.

Ghiglione et al. PNAS Early Edition | 5 of 6

MICRO

BIOLO

GY

ence between OTUs (maximum differences at the poles), aremore likely to be sympatric.Our bipolar comparisons highlighted global patterns of bio-

geography and diversity in pelagic bacterioplankton. A notablelack of correlation between latitude and richness warrants ad-ditional attention, and further research is needed to address theecological and evolutionary processes underlying these patterns.Our comprehensive global dataset, encompassing diverse oce-anic regions, suggests that surface communities are driven byenvironmental selection, whereas deep communities are moreconstrained by historical events and connected through oceaniccirculation, providing evidence for biogeographically definedcommunities in the global ocean.

Materials and MethodsAll samples (Fig. 1) were part of the International Census of Marine Microbes(ICoMM), which developed sample and metadata submission, sample pro-cessing, and data analysis pipelines, ensuring similar treatment for all sam-ples. The ICoMM hypervariable V6 region SSU rRNA pyrotag dataset andgeospatial parameters are available at http://icomm.mbl.edu/microbis andare reported in Table S1. All Southern Ocean samples except the AmundsenSea vertical profile were coordinated by the Census of Antarctic Marine Life(CAML) with the goal of representing spatial, temporal, and latitudinal var-iability across four subregions of the Southern Ocean including the Weddelland Ross Seas, the Antarctic Peninsula, and Kerguelen Islands. Samples col-lected over the annual cycle in the Antarctic Peninsula and Kerguelen Islandshave been described recently (34). The Amundsen Sea profile was collectedon the same cruise as described recently (35). We classified samples based onwhether they were from the Southern or the Arctic Ocean, coastal or open

ocean (limit of 200 km from the coast), surface (<40 m) or intermediate anddeep waters (>200 m), summer or winter. The Arctic Ocean samples includedin this study have been reported previously (30, 36, 37).

Details concerning the methodological approach and data analysis pipe-line are described in SI Materials and Methods. In brief, the approach in-cluded data resampling, and a pipeline using Mothur (38) including multiplealignment, preclustering, distance matrix calculation, and clustering to de-fine OTUs at similarity thresholds ranging from 87% to 100% sequenceidentity. Details of the statistical approaches used are also described in thesupplementary information and included hierarchical agglomerative clus-tering of Bray–Curtis similarities, SIMPROF, one-way ANOSIM, similaritypercentage (SIMPER), UniFrac, canonical correspondence analysis (CCA), andSpearman rank correlations. All sequence tags used are publicly available athttp://vamps.mbl.edu as trimmed FASTA sequences (used in our analysis) oras ICoMM processed datasets.

ACKNOWLEDGMENTS. We thank the members of field teams, shipboardcrews, and logistics support personnel from all national polar programs in-volved in sample collection, without whom this study would not have beenpossible. The Census of Antarctic Marine Life, funded by the Sloan Foun-dation, facilitated this collaboration. Pyrosequencing was provided by theInternational Census of Marine Microbes (ICoMM) with financial supportfrom a W. M. Keck Foundation award to the Marine Biological Laboratory inWoods Hole. Funding to support sample collection was provided by theInstitut Français pour la Recherche et la Technologie Polaires (J.-F.G.); theSpanish Ministry of Education and Science (C.P.-A.); the New Zealand In-ternational Polar Year-Census of Antarctic Marine Life Project [Phases 1(So001IPY) and 2 (IPY2007-01); to E.W.M.); the Natural Sciences and Engi-neering Council (NSERC) of Canada (C.L.); National Science FoundationGrants OPP-0124733 (to D.L.K.), ANT-0632389 (to A.E.M.), and ANT-0741409(to P.L.Y.); and the Swedish Polar Research Secretariat (S.B.).

1. Nemergut DR, et al. (2011) Global patterns in the biogeography of bacterial taxa.Environ Microbiol 13:135–144.

2. Zinger L, et al. (2011) Global patterns of bacterial beta-diversity in seafloor andseawater ecosystems. PLoS ONE 6:e24570.

3. Siegert MJ, et al. (2008) Recent advances in understanding Antarctic climate evolu-tion. Antarct Sci 4:313–325.

4. Karl DM (1993) Antarctic Microbiology, ed Friedmann EI (Wiley, New York), pp 1–63.5. Wheeler PA, et al. (1996) Active cycling of organic carbon in the central Arctic Ocean.

Nature 380:697–699.6. Cochran JR, Edwards MH, Coakley BJ (2006) Morphology and structure of the Lo-

monosov Ridge, Arctic Ocean. Geochem Geophys Geosyst 7:Q05019.7. Verity PG, et al. (2007) Current understanding of Phaeocystis ecology and bio-

geochemistry, and perspectives for future research. Biogeochemistry 83:311–330.8. Darling KF, Kucera M, Wade CM (2007) Global molecular phylogeography reveals

persistent Arctic circumpolar isolation in a marine planktonic protist. Proc Natl AcadSci USA 104:5002–5007.

9. Staley JT, Gosink JJ (1999) Poles apart: Biodiversity and biogeography of sea icebacteria. Annu Rev Microbiol 53:189–215.

10. Zeng Y, Zheng T, Yu Y, Chen B, He J (2010) Relationships between Arctic and AntarcticShewanella strains evaluated by a poly- phasic taxonomic approach. Polar Biol 33:531–541.

11. Bano N, Ruffin S, Ransom B, Hollibaugh JT (2004) Phylogenetic composition of ArcticOcean archaeal assemblages and comparison with Antarctic assemblages. Appl En-viron Microbiol 70:781–789.

12. Baldwin AJ, et al. (2005) Microbial diversity in a Pacific Ocean transect from the Arcticto Antarctic circles. Aquat Microb Ecol 41:91–102.

13. Brinkmeyer R, et al. (2003) Diversity and structure of bacterial communities in Arcticversus Antarctic pack ice. Appl Environ Microbiol 69:6610–6619.

14. Pommier T, Pinhassi J, Hagström Å (2005) Biogeographic analysis of ribosomal RNAclusters from marine bacterioplankton. Aquat Microb Ecol 41:79–89.

15. Amaral-Zettler L, et al. (2010) Life in the World’s Oceans: Diversity, Distribution, andAbundance, ed McIntyre AD (Wiley-Blackwell, Oxford), pp 223–245.

16. Huse SM, Huber JA, Morrison HG, Sogin ML, Welch DM (2007) Accuracy and quality ofmassively parallel DNA pyrosequencing. Genome Biol 8:R143.

17. Pawlowski J, et al. (2007) Bipolar gene flow in deep-sea benthic foraminifera. MolEcol 16:4089–4096.

18. Montresor M, Lovejoy C, Orsini L, Procaccini G, Roy S (2003) Bipolar distribution of thecyst-forming dinoflagellate Polarella glacialis. Polar Biol 26:186–194.

19. Gosink JJ, Herwig RP, Staley JT (1997) Octadecabacter arcticus gen. nov., sp. nov.,nonpigmented, psychrophilic gas vacuolate bacteria from polar sea ice and water.Syst Appl Microbiol 20:356–365.

20. Crump BC, et al. (2009) Circumpolar synchrony in big river bacterioplankton. Proc NatlAcad Sci USA 106:21208–21212.

21. Ghiglione JF, Larcher M, Lebaron P (2005) Spatial and temporal scales of variation inbacterioplankton community structure in the NW Mediterranean Sea. Aquat MicrobEcol 40:229–240.

22. Lee SH, Fuhrman JA (1991) Spatial and temporal variation of natural bacter-

ioplankton assemblages studied by total genomic DNA cross-hybridization. Limnol

Oceanogr 36:1277–1287.23. Field KG, et al. (1997) Diversity and depth-specific distribution of SAR11 cluster rRNA

genes from marine planktonic bacteria. Appl Environ Microbiol 63:63–70.24. Murray AE, et al. (1998) Seasonal and spatial variability of bacterial and archaeal

assemblages in the coastal waters near Anvers Island, Antarctica. Appl Environ Mi-

crobiol 64:2585–2595.25. Fuhrman JA, et al. (2008) A latitudinal diversity gradient in planktonic marine bac-

teria. Proc Natl Acad Sci USA 105:7774–7778.26. Aagaard K, Swift JH, Carmack EC (1985) Thermohaline circulation in the Arctic

Mediterranean Seas. J Geophys Res 90:4833–4846.27. Galand PE, Potvin M, Casamayor EO, Lovejoy C (2010) Hydrography shapes bacterial

biogeography of the deep Arctic Ocean. ISME J 4:564–576.28. Beszczynska-Möller A, Woodgate RA, Lee C, Melling H, Karcher M (2011) A synthesis

of exchanges through the main Oceanic gateways to the Arctic Ocean. Oceanography

(Wash DC) 24:82–99.29. Gordon DA, Giovannoni SJ (1996) Detection of stratified microbial populations re-

lated to Chlorobium and Fibrobacter species in the Atlantic and Pacific oceans. Appl

Environ Microbiol 62:1171–1177.30. Barberán A, Fernández-Guerra A, Auguet JC, Galand PE, Casamayor EO (2011) Phy-

logenetic ecology of widespread uncultured clades of the Kingdom Euryarchaeota.

Mol Ecol 20:1988–1996.31. Pommier T, Douzery EJ, Mouillot D (2012) Environment drives high phylogenetic

turnover among oceanic bacterial communities. Biol Lett 8:562–566.32. Johnson MTJ, Stinchcombe JR (2007) An emerging synthesis between community

ecology and evolutionary biology. Trends Ecol Evol 22:250–257.33. Whitaker RJ (2006) Allopatric origins of microbial species. Philos Trans R Soc Lond B

Biol Sci 361:1975–1984.34. Ghiglione JF, Murray AE (2012) Pronounced summer to winter differences and higher

wintertime richness in coastal Antarctic marine bacterioplankton. Environ Microbiol

14:617–629.35. Alonso-Sáez L, Andersson A, Heinrich F, Bertilsson S (2011) High archaeal diversity in

Antarctic circumpolar deep waters. Environ Microbiol 3:689–697.36. Galand PE, Casamayor EO, Kirchman DL, Potvin M, Lovejoy C (2009) Unique archaeal

assemblages in the Arctic Ocean unveiled by massively parallel tag sequencing. ISME J

3:860–869.37. Kirchman DL, Cottrell MT, Lovejoy C (2010) The structure of bacterial communities in

the western Arctic Ocean as revealed by pyrosequencing of 16S rRNA genes. Environ

Microbiol 12:1132–1143.38. Schloss PD, et al. (2009) Introducing mothur: Open-source, platform-independent,

community-supported software for describing and comparing microbial communities.

Appl Environ Microbiol 75:7537–7541.

6 of 6 | www.pnas.org/cgi/doi/10.1073/pnas.1208160109 Ghiglione et al.

Supporting InformationGhiglione et al. 10.1073/pnas.1208160109SI Materials and MethodsData Analysis Pipeline. The hypervariable V6 region of the SSUrRNA gene in bacteria was amplified using a combination of 967F(forward) and 1046R (reverse) primers under conditions de-scribed previously (1). Amplicons were sequenced using py-rosequencing technology (454 Life Sciences; GS20) at theMarine Biological Laboratory. Random sequencing errors wereremoved as described previously (2). To normalize the numberof tags sequenced between samples and datasets and locations,tags were randomly resampled to the sample with the fewest tagsusing Daisychopper version 0.6 (3). Analyses of tag sequenceswere based on the frequency of resampled tags analyzed inMothur (4) following a work flow that included multiple align-ment to Silva database reference alignment (June 29, 2010),error-minimization using a “preclustering” step, and then a dis-tance matrix was generated as input for “cluster,” from whichoutput files were created. Clusters based on a distance of 0.03(≥97% identity) were used in all downstream analyses (thesehave up to 2 mismatches per 60-bp sequence). In addition,Mothur was used to define OTUs at similarity thresholds rangingfrom 87 to 100%. The number of remaining OTUs diminishedwith the decreasing SSU rRNA sequence similarity thresholdused to define OTUs, and the percentage of remaining OTUsbetween one threshold and the next was plotted for each simi-larity value. Differences in the proportion of remaining OTUsbetween samples were used as a metric for comparison of thephylogenetic structure between communities (5). Taxonomicidentification of the sequence reads (tags) followed the approachby Sogin et al. (6) and Huse et al. (7). Representative sequencesfor clustered tags originating from a global reference dataset atthe 0.03 distance level are available at http://vamps.mbl.edu.

Similarity-Based Cluster and Statistical Analyses. Similarity-basedcluster and statistical analyses were performed using the un-weighted pair group method with arithmetic mean (UPGMA)(Primer version 6 software). A similarity profile test (SIMPROF)

(Primer version 6) was performed on a null hypothesis that aspecific subcluster can be recreated by permuting the entry speciesand samples. The significant branch (SIMPROF, P < 0.05) wasused as a prerequisite for defining bacterial clusters. One-wayanalysis of similarity (ANOSIM) (Primer version 6) was per-formed on the same distance matrix to test the null hypothesisthat there was no difference between bacterial communities ofdifferent clusters. Similarity percentage (SIMPER) (8) was usedto determine which individual OTUs contributed most to thedissimilarity between groups of samples.The resulting Bray–Curtis distance matrix was used to build a

UPGMA tree of sample relationships. This resulting tree wasimported in the web version of UniFrac (9), together with re-spective groups according to depth, distance from shore, andregions. UniFrac and P test significance were performed on allsamples together and pairs of samples.Relationships between bacterioplankton communities and

environmental parameters were assessed using a direct gradientapproach, canonical correspondence analysis (CCA) (CANOCOversion 4.5) according to ref. 10. Spearman rank pairwise cor-relations between the environmental variables (depth, tempera-ture, salinity, chlorophyll, latitude, and longitude; Table S1)helped to determine their significance. To statistically evaluatethe significance of the first canonical axis and of all canonicalaxes together, we used Monte Carlo permutation full model testwith 199 unrestricted permutations. Significant variables werechosen using a forward-selection procedure and 999 permuta-tions. Furthermore, we specifically examined the relationshipbetween latitude and OTU richness based on the Chao1 richnessestimator calculated using the application in the Mothur packageof executable programs. The analysis was limited to surfacewater samples at depths of 0–40 m (only two samples were col-lected at depths below 11 m) and deep samples collected atdepths below 200 m. The Spearman Rank correlation (and sig-nificance) was calculated for latitude vs. Chao1 datasets in the“surface” and “deep.”

1. Huber JA, et al. (2007) Microbial population structures in the deep marine biosphere.Science 318:97–100.

2. Huse SM, Huber JA, Morrison HG, Sogin ML, Welch DM (2007) Accuracy and quality ofmassively parallel DNA pyrosequencing. Genome Biol 8:R143.

3. Gilbert JA, et al. (2009) The seasonal structure of microbial communities in theWestern English Channel. Environ Microbiol 11:3132–3139.

4. Schloss PD, et al. (2009) Introducing mothur: Open-source, platform-independent,community-supported software for describing and comparing microbial communities.Appl Environ Microbiol 75:7537–7541.

5. Barberán A, Fernández-Guerra A, Auguet JC, Galand PE, Casamayor EO (2011)Phylogenetic ecology of widespread uncultured clades of the Kingdom Euryarchaeota.Mol Ecol 20:1988–1996.

6. Sogin ML, et al. (2006) Microbial diversity in the deep sea and the underexplored“rare biosphere”. Proc Natl Acad Sci USA 103:12115–12120.

7. Huse SM, Welch DM, Morrison HG, Sogin ML (2010) Ironing out the wrinkles inthe rare biosphere through improved OTU clustering. Environ Microbiol 12:1889–1898.

8. Clarke KR, Warwick RW (2001) Change in Marine Communities: An Approach toStatistical Analysis and Interpretation (PRIMER-E, Plymouth, UK), 2nd Ed, pp 1–72.

9. Lozupone C, Hamady M, Knight R (2006) UniFrac—an online tool for comparingmicrobial community diversity in a phylogenetic context. BMC Bioinformatics 7:371.

10. Ghiglione JF, Larcher M, Lebaron P (2005) Spatial and temporal scales of variation inbacterioplankton community structure in the NW Mediterranean Sea. Aquat MicrobEcol 40:229–240.

Ghiglione et al. www.pnas.org/cgi/content/short/1208160109 1 of 7

Fig. S1. Relationship between latitude and OTU richness as estimated by Chao1 for surface (0–40 m) (A) and deep (>200 m) (B) 16S rRNA gene tag datasets.The absolute value of latitude is plotted with the Southern and Northern hemisphere datasets indicated with circles (open circles for summer samples, andclosed circles for winter samples) and squares, respectively.

Ghiglione et al. www.pnas.org/cgi/content/short/1208160109 2 of 7

AO2448 (70,4%)

AO & SO1029

SO3661 (78,0%)

AO&SO739

SO606 (45%)

(A) SURFACE

(B) DEEP

0

20

40

60

80

100

0

20

40

60

80

100

0

20

40

60

80

100

0

20

40

60

80

100

0

20

40

60

80

100

0

20

40

60

80

100

AO4116 (84,8%)

Fig. S2. Venn diagrams representing surface (A) and deep (B) OTUs found in Southern Ocean (SO) and Arctic Ocean (AO) bacterial datasets. Number of OTUsand their relative percentage to the total number of OTU (in brackets) are indicated in bold, and the corresponding number of tag sequences are indicated inbrackets. Histograms of corresponding taxonomic composition are grouped at the class level for unique and shared OTUs between polar communities.

Ghiglione et al. www.pnas.org/cgi/content/short/1208160109 3 of 7

Surface ocean OTUs common to each pair of data sets

SO AO MLSO 3264 1029 817AO - 2002 666ML - 1104

Percents of surface ocean OTUs common to each pair of data sets

SO AO MLSO 69.6 14.8 5.1AO - 61.1 4.5ML - 91.3

Deep ocean OTUs common to each pair of data sets

SO AO MLSO 528 739 627AO - 2920 1744ML - - 7371

Percents of deep ocean OTUs common to each pair of data sets

SO AO MLSO 39.2 11.9 5.9AO - 60.1 12.4ML - - 80.2

Fig. S3. OTU groupings when all bacterial samples [Southern Ocean (SO), Arctic Ocean (AO), and midlatitude (ML)] were compared in surface and deep oceandatasets. Note that 420 OTUs (2.1%) were common to all three datasets in surface waters, and 548 OTUs (3.6%) were common in deep waters.

Ghiglione et al. www.pnas.org/cgi/content/short/1208160109 4 of 7

Table

S1.

Stationan

dsample

metad

ata,

categories,an

den

vironmen

talparam

eters

Sample

nam

eStationID

(ICoMM

form

at)

withsamplin

gdate

Dep

th(m

)La

titude

Longitude

Ocean

region

Ocean

subregion

Total

sequen

ces

Total

OTU

s(0.03)

Salin

ity

(PSU

)Te

mperature

(°C)

Chlorophyll

(μg×L−

1)

AO_B

S1_C

2ACB_0

001_

2007

_07_

132

71.45

−15

6.06

Arctic

Bea

ufort

Sea

15,600

786

20.0

2.5

0.33

AO_C

S2_C

2ACB_0

002_

2007

_07_

112

71.43

−15

6.86

Arctic

ChukchiSe

a9,10

746

232

.04.6

0.50

AO_C

S3_C

2ACB_0

003_

2008

_01_

262

71.35

−15

6.68

Arctic

ChukchiSe

a17

,516

387

35.0

−1.8

0.07

AO_C

S4_C

2ACB_0

004_

2008

_01_

302

71.35

−15

6.68

Arctic

ChukchiSe

a19

,118

666

35.0

−2.0

0.06

AO_C

S5_O

10ACB_0

005_

2002

_07_

2910

72.32

−15

1.98

Arctic

ChukchiSe

a19

,737

517

29.4

−1.5

0.57

AO_C

S6_O

10ACB_0

006_

2002

_08_

0610

72.24

−15

9.34

Arctic

ChukchiSe

a14

,575

472

29.5

−1.1

0.22

AO_B

S7_O

10ACB_0

007_

2004

_07_

3010

71.54

−15

0.89

Arctic

Bea

ufort

Sea

17,551

465

29.9

5.2

0.64

AO_B

S8_O

40ACB_0

008_

2004

_08_

2140

72.42

−15

2.02

Arctic

Bea

ufort

Sea

14,113

427

30.3

−1.2

0.27

AO_F

B13

_C3*

ACB_0

013_

2004

_01_

173

70.04

−12

6.30

Arctic

Fran

klin

Bay

12,389

664

29.8

−1.6

0.04

AO_F

B15

_C3

ACB_0

015_

2004

_07_

163

70.05

−12

6.31

Arctic

Fran

klin

Bay

50,884

527

29.0

4.0

0.61

AO_C

S16_

C2*

ACB_0

016_

2008

_01_

282

71.35

−15

6.68

Arctic

ChukchiSe

a19

,084

799

35.0

−1.8

0.06

SO_A

S1_C

10ASA

_000

1_20

07_1

2_18

10−73

.94

−11

5.68

Antarctic

AmundsenSe

a11

,411

182

33.9

−2.0

8.16

SO_A

S10_

C25

0ASA

_001

0_20

07_1

2_18

250

−73

.94

−11

5.68

Antarctic

AmundsenSe

a15

,636

364

34.0

−1.8

1.29

SO_A

S13_

C79

0ASA

_001

3_20

07_1

2_18

790

−73

.96

−11

5.86

Antarctic

AmundsenSe

a23

,704

802

34.5

0.4

NA

SO_A

S14_

C50

0ASA

_001

4_20

07_1

2_18

500

−73

.96

−11

5.86

Antarctic

AmundsenSe

a27

,983

717

34.1

−1.6

NA

SO_A

P1_C

6CAM_0

001_

2002

_01_

176

−64

.77

−64

.05

Antarctic

AntarcticPe

ninsula

23,733

442

33.6

1.6

NA

SO_A

P2_C

6*CAM_0

002_

2002

_07_

176

−64

.77

−64

.05

Antarctic

AntarcticPe

ninsula

10,798

826

33.8

−0.7

0.04

SO_A

P3_C

6*CAM_0

003_

2002

_08_

206

−64

.77

−64

.05

Antarctic

AntarcticPe

ninsula

12,662

962

33.8

−0.37

0.07

SO_A

P4_C

6CAM_0

004_

2006

_02_

286

−64

.77

−64

.07

Antarctic

AntarcticPe

ninsula

22,147

356

33.4

2.2

2.97

SO_K

I5_C

3*CAM_0

005_

2007

_07_

233

−49

.37

−70

.20

Antarctic

Kerguelen

Islands

38,796

743

32.8

1.9

0.16

SO_K

I6_C

3CAM_0

006_

2007

_10_

093

−49

.37

−70

.20

Antarctic

Kerguelen

Islands

32,147

563

33.3

3.2

5.04

SO_K

I7_C

3CAM_0

007_

2007

_02_

073

−49

.37

−70

.20

Antarctic

Kerguelen

Islands

18,711

562

33.8

6.9

0.35

SO_K

I8_C

3*CAM_0

008_

2007

_05_

043

−49

.37

−70

.20

Antarctic

Kerguelen

Islands

24,255

711

33.4

4.9

0.29

SO_W

S9_O

11CAM_0

009_

2000

_03_

2311

−55

.13

−2.89

Antarctic

Wed

dellSe

a20

,673

545

NA

0.9

0.27

SO_W

S10_

O11

CAM_0

010_

2000

_03_

2511

−59

.32

−0.40

Antarctic

Wed

dellSe

a29

,214

364

34.2

0.3

0.04

SO_W

S11_

O11

CAM_0

011_

2000

_03_

2611

−67

.02

−5.47

Antarctic

Wed

dellSe

a19

,583

412

NA

−0.8

0.05

SO_W

S12_

C11

CAM_0

012_

2000

_03_

2911

−71

.18

−12

.46

Antarctic

Wed

dellSe

a21

,290

496

34.2

−1.8

0.05

SO_R

S13_

O10

CAM_0

013_

2008

_02_

0410

−59

.40

175.56

Antarctic

Ross

Sea

15,506

429

33.9

4.3

0.25

SO_R

S14_

O10

CAM_0

014_

2008

_02_

0610

−65

.01

179.97

Antarctic

Ross

Sea

17,998

413

33.6

1.0

0.24

SO_R

S15_

O10

CAM_0

015_

2008

_02_

2910

−69

.43

−17

9.09

Antarctic

Ross

Sea

23,149

441

33.5

−1.8

0.38

SO_R

S16_

C10

CAM_0

016_

2008

_02_

1310

−75

.66

169.77

Antarctic

Ross

Sea

19,471

461

NA

−0.8

0.69

AO_B

B1_

O10

00DAO_0

001_

2007

_07_

121,00

068

.83

−61

.76

Arctic

BaffinBay

27,165

720

34.5

0.9

0.00

AO_B

S3_O

1000

DAO_0

003_

2007

_08_

041,00

071

.96

−15

0.23

Arctic

Bea

ufort

Sea

18,873

916

34.9

0.0

NA

AO_B

S4_O

400

DAO_0

004_

2007

_08_

0440

071

.96

−15

0.23

Arctic

Bea

ufort

Sea

24,725

806

34.8

0.8

0.01

AO_B

S5_O

1000

DAO_0

005_

2007

_08_

121,00

079

.99

−14

9.99

Arctic

Bea

ufort

Sea

17,654

871

34.9

0.0

0.00

AO_B

S7_O

400

DAO_0

007_

2007

_08_

1240

079

.99

−14

9.99

Arctic

Bea

ufort

Sea

31,400

821

34.8

0.1

0.00

AO_B

S9_O

1000

DAO_0

009_

2007

_08_

151,00

077

.00

−14

0.19

Arctic

Bea

ufort

Sea

14,217

1,19

034

.90.0

NA

AO_B

S10_

O40

0DAO_0

010_

2007

_08_

1540

077

.00

−14

0.19

Arctic

Bea

ufort

Sea

20,328

1,02

634

.80.9

0.00

AO_B

S11_

O90

0DAO_0

011_

2007

_08_

2090

075

.84

−12

8.64

Arctic

Bea

ufort

Sea

21,373

1,06

234

.90.1

NA

AO_B

S12_

O10

00DAO_0

012_

2007

_08_

231,00

073

.97

−14

0.09

Arctic

Bea

ufort

Sea

22,612

1,10

934

.90.0

0.00

AO_B

S13_

O40

0DAO_0

013_

2007

_08_

2340

073

.97

−14

0.09

Arctic

Bea

ufort

Sea

16,487

1,00

032

.80.7

0.00

AO_E

S14_

O10

00DAO_0

014_

2007

_09_

181,00

077

.75

126.00

Arctic

East

SiberianSe

a36

,728

894

34.9

0.0

0.00

AO_E

S15_

O12

00DAO_0

015_

2007

_09_

211,20

079

.94

142.39

Arctic

East

SiberianSe

a26

,671

882

34.9

−0.3

0.00

AO_E

S16_

O25

0DAO_0

016_

2007

_09_

2125

079

.94

142.39

Arctic

East

SiberianSe

a15

,704

967

34.8

1.4

0.00

Ghiglione et al. www.pnas.org/cgi/content/short/1208160109 5 of 7

Table

S1.

Cont.

Sample

nam

eStationID

(ICoMM

form

at)

withsamplin

gdate

Dep

th(m

)La

titude

Longitude

Ocean

region

Ocean

subregion

Total

sequen

ces

Total

OTU

s(0.03)

Salin

ity

(PSU

)Te

mperature

(°C)

Chlorophyll

(μg×L−

1)

NA11

2_O40

00KCK_N

ADP_

Bv6

;112

R_4

121m

4,00

050

.40

−25

.00

NorthAtlan

tic

NorthAtlan

ticDee

pWater

15,497

1,35

634

.92.3

NA

NA11

5_O55

0KCK_N

ADP_

Bv6

;115

R_5

50m

550

50.40

−25

.00

NorthAtlan

tic

NorthAtlan

ticDee

pWater

16,334

1,02

135

.17.0

NA

NA13

7_O17

00KCK_N

ADP_

Bv6

;137

_171

0m1,70

060

.90

38.52

NorthAtlan

tic

NorthAtlan

ticDee

pWater

13,759

946

34.9

3.0

NA

NA13

8_O70

0KCK_N

ADP_

Bv6

;138

–71

0m70

060

.90

38.52

NorthAtlan

tic

NorthAtlan

ticDee

pWater

12,980

969

34.9

3.5

NA

NA53

_O14

00KCK_N

ADP_

Bv6

;53R

_140

0m1,40

058

.30

29.13

NorthAtlan

tic

NorthAtlan

ticDee

pW

ater

12,763

1,07

8NA

NA

NA

NA55

_O50

0KCK_N

ADP_

Bv6

;55R

_500

m50

058

.30

29.13

NorthAtlan

tic

NorthAtlan

ticDee

pWater

9,90

31,21

935

.17.1

NA

MS2

_C5

BMO_0

002_

2007

_09_

205

41.65

2.80

Med

iterranea

nNW

Med

iterranea

nSe

a18

,361

580

37.9

23.3

0.09

MS3

_C5

BMO_0

003_

2007

_09_

215

41.40

2.80

Med

iterranea

nNW

Med

iterranea

nSe

a22

,253

326

37.8

23.8

0.07

MS4

_O5

BMO_0

004_

2007

_09_

215

41.15

2.82

Med

iterranea

nNW

Med

iterranea

nSe

a17

,311

360

37.9

23.8

0.08

MS5

_O5

BMO_0

005_

2007

_09_

225

40.91

2.85

Med

iterranea

nNW

Med

iterranea

nSe

a22

,951

456

37.9

21.0

0.10

MS6

_O5

BMO_0

006_

2007

_09_

235

40.65

2.85

Med

iterranea

nNW

Med

iterranea

nSe

a21

,884

346

37.3

24.3

0.08

MS9

_O50

0BMO_0

009_

2007

_09_

2250

040

.66

2.86

Med

iterranea

nNW

Med

iterranea

nSe

a23

,347

711

38.516

13.185

NA

MS1

0_O20

00BMO_0

010_

2007

_09_

222,00

040

.66

2.86

Med

iterranea

nNW

Med

iterranea

nSe

a14

,527

987

38.479

13.213

NA

MS1

1_C50

0BMO_0

011_

2007

_09_

2250

040

.65

2.85

Med

iterranea

nNW

Med

iterranea

nSe

a12

,585

764

38.516

13.185

NA

NA_1

_O47

AOT_

0001

_200

8_11

_07

4742

.85

−11

.64

NorthAtlan

tic

NorthAtlan

tic

26,972

890

35.92

16NA

NA_2

_O11

00AOT_

0002

_200

8_11

_09

1,10

037

.13

−13

.36

NorthAtlan

tic

NorthAtlan

tic

18,090

1,33

236

.017

10.926

5NA

NA_3

_O73

AOT_

0003

_200

8_11

_09

7337

.13

−13

.36

NorthAtlan

tic

NorthAtlan

tic

24,795

1,14

036

.069

715

.795

4NA

NA_4

_O89

AOT_

0004

_200

8_11

_14

8922

.51

−20

.50

NorthAtlan

tic

NorthAtlan

tic

25,787

951

36.847

822

.755

6NA

NA_5

_O7

AOT_

0005

_200

8_11

_16

714

.76

−30

.00

NorthAtlan

tic

NorthAtlan

tic

21,824

708

35.65

26.473

0.17

SA_6

_O28

*AOT_

0006

_200

8_11

_29

28−23

.71

8.52

South

Atlan

tic

South

Atlan

tic

18,532

702

35.567

318

.066

70.18

NA_7

_O7

AOT_

0007

_200

8_11

_17

744

.70

−20

.36

NorthAtlan

tic

NorthAtlan

tic

27,567

847

35.44

15.6

0.32

NA_8

_O10

0AOT_

0008

_200

8_11

_17

100

10.63

−20

.13

NorthAtlan

tic

NorthAtlan

tic

11,722

1,16

735

.573

915

.288

7NA

NA_9

_O48

AOT_

0009

_200

8_11

_17

4810

.63

−20

.13

NorthAtlan

tic

NorthAtlan

tic

46,751

945

35.972

420

.434

8NA

NA_1

0_O11

AOT_

0010

_200

8_11

_17

1110

.37

−20

.06

NorthAtlan

tic

NorthAtlan

tic

20,884

723

35.09

28.214

0.39

NA_1

2_O13

00AOT_

0012

_200

8_11

_17

1,30

010

.37

−20

.06

NorthAtlan

tic

NorthAtlan

tic

16,117

1,15

034

.955

4.60

8NA

NA_1

3_O7

AOT_

0013

_200

8_11

_18

78.50

−19

.50

NorthAtlan

tic

NorthAtlan

tic

23,799

714

34.48

28.608

0.21

SA_1

4_O46

00*

AOT_

0014

_200

8_11

_29

4,60

0−23

.71

8.52

South

Atlan

tic

South

Atlan

tic

10,151

949

35.6

3.44

7NA

SA_1

6_O48

*AOT_

0016

_200

8_11

_27

48−17

.74

3.13

South

Atlan

tic

South

Atlan

tic

17,126

545

35.88

18.701

NA

AZ_

1_O0

AW

P_00

01_2

007_

08_2

30

36.05

−28

.35

Azo

res

Azo

res

17,185

727

36.73

23.6

0.08

AZ_

3_O0

AW

P_00

03_2

007_

08_2

50

34.21

−32

.12

Azo

res

Azo

res

22,064

746

36.932

25.3

0.07

AZ_

6_O0

AW

P_00

06_2

007_

08_2

50

34.21

−32

.12

Azo

res

Azo

res

20,528

623

36.932

25.3

0.07

AZ_

7_O0

AW

P_00

07_2

007_

08_2

50

35.48

−32

.43

Azo

res

Azo

res

17,419

614

36.863

24.6

0.08

AZ_

8_O0

AW

P_00

08_2

007_

08_2

60

37.48

−30

.57

Azo

res

Azo

res

25,041

596

36.172

23.1

0.09

AZ_

9_O36

60AW

P_00

09_2

007_

06_1

13,66

037

.34

−18

.88

Azo

res

Azo

res

25,840

1,04

434

.92

2.56

NA

AZ_

10_O

2100

AW

P_00

10_2

007_

06_1

12,10

037

.34

−18

.88

Azo

res

Azo

res

24,343

972

353.69

NA

AZ_

11_O

1200

AW

P_00

11_2

007_

06_1

11,20

037

.34

−18

.88

Azo

res

Azo

res

20,825

1,23

835

.84

9.11

NA

AZ_

12_O

800

AW

P_00

12_2

007_

06_1

180

037

.34

−18

.88

Azo

res

Azo

res

23,238

1,30

135

.67

10.33

NA

AZ_

13_O

100

AW

P_00

13_2

007_

06_1

110

037

.34

−18

.88

Azo

res

Azo

res

24,435

842

36.19

16.37

NA

AZ_

14_O

0AW

P_00

14_2

007_

06_1

10

37.34

−18

.88

Azo

res

Azo

res

25,370

585

36.4

18.54

0.12

AZ_

15_O

800

AW

P_00

15_2

007_

06_0

280

031

.86

−27

.88

Azo

res

Azo

res

21,776

1,13

136

.19

10NA

AZ_

16_O

100

AW

P_00

16_2

007_

06_0

210

031

.86

−27

.88

Azo

res

Azo

res

27,904

1,39

837

18.54

NA

SP_2

_O11

0KNX_0

002_

2006

_12_

2411

0−26

.05

−15

6.89

South

Pacific

South

Pacific

26,953

1,15

435

.57

18.13

0.39

SP3_

O12

0KNX_0

003_

2006

_12_

2712

0−27

.94

−14

8.59

South

Pacific

South

Pacific

23,273

1,11

835

.518

.50.36

SP_4

_O15

0KNX_0

004_

2006

_12_

3015

0−26

.48

−13

7.94

South

Pacific

South

Pacific

18,093

1,19

835

.519

.37

0.22

Ghiglione et al. www.pnas.org/cgi/content/short/1208160109 6 of 7

Table

S1.

Cont.

Sample

nam

eStationID

(ICoMM

form

at)

withsamplin

gdate

Dep

th(m

)La

titude

Longitude

Ocean

region

Ocean

subregion

Total

sequen

ces

Total

OTU

s(0.03)

Salin

ity

(PSU

)Te

mperature

(°C)

Chlorophyll

(μg×L−

1)

SP_5

_O14

5KNX_0

005_

2007

_01_

0114

5−28

.45

−13

1.39

South

Pacific

South

Pacific

24,661

1,21

735

.39

18.1

0.35

SP_6

_O20

0KNX_0

006_

2007

_01_

0720

0−27

.74

−11

7.62

South

Pacific

South

Pacific

30,918

1,27

735

.518

.76

0.27

SP_7

_O11

0KNX_0

007_

2007

_01_

1111

0−38

.06

−13

3.09

South

Pacific

South

Pacific

21,932

891

34.3

11.5

0.30

SP_8

_O10

0KNX_0

008_

2007

_01_

1310

0−39

.31

−13

9.80

South

Pacific

South

Pacific

16,099

792

34.36

11.2

0.40

Term

sforsample

nam

eweredesignated

byab

breviationsderived

from

theocean

region,subregion,a

nddep

th.E

nvironmen

talp

aram

etersderived

from

remotely

senseddataareindicated

inboldface

italics

(selecttemperature

andch

lorophyllva

lues).NA,notav

ailable.

Asterisks

(*)indicatesamplesco

llected

inthewinter.

Ghiglione et al. www.pnas.org/cgi/content/short/1208160109 7 of 7