Spatial variability overwhelms seasonal patterns in bacterioplankton communities across a river to ocean gradient

ORIGINAL ARTICLE

Spatial variability overwhelms seasonal patternsin bacterioplankton communities across a riverto ocean gradient

Caroline S Fortunato1, Lydie Herfort2, Peter Zuber2, Antonio M Baptista2

and Byron C Crump1

1Horn Point Laboratory, University of Maryland Center for Environmental Science, Cambridge, MD, USAand 2Center for Coastal Margin Observation & Prediction, Oregon Health & Sciences University, Beaverton,OR, USA

Few studies of microbial biogeography address variability across both multiple habitats andmultiple seasons. Here we examine the spatial and temporal variability of bacterioplanktoncommunity composition of the Columbia River coastal margin using 16S amplicon pyrosequencingof 300 water samples collected in 2007 and 2008. Communities separated into seven groups(ANOSIM, Po0.001): river, estuary, plume, epipelagic, mesopelagic, shelf bottom (deptho350 m)and slope bottom (depth4850 m). The ordination of these samples was correlated with salinity(q¼�0.83) and depth (q¼�0.62). Temporal patterns were obscured by spatial variability among thecoastal environments, and could only be detected within individual groups. Thus, structuringenvironmental factors (for example, salinity, depth) dominate over seasonal changes in determiningcommunity composition. Seasonal variability was detected across an annual cycle in the river,estuary and plume where communities separated into two groups, early year (April–July) and lateyear (August–Nov), demonstrating annual reassembly of communities over time. Determining boththe spatial and temporal variability of bacterioplankton communities provides a framework formodeling these communities across environmental gradients from river to deep ocean.The ISME Journal (2012) 6, 554–563; doi:10.1038/ismej.2011.135; published online 20 October 2011Subject Category: microbial population and community ecologyKeywords: coastal ocean; Columbia River; gradient; spatial variability; 16S amplicon pyrosequencing;temporal variability

Introduction

Over the past few decades, aquatic microbialcommunities have been shown to be abundant,deeply diverse, and variable across space and time.Yet several recent studies demonstrate repeatableand predictable patterns in the composition ofthese communities. Spatial variability in aquaticmicrobial communities has been explored on scalesthat range from millimeters (Long and Azam, 2001)to kilometers (Hewson et al., 2006) to global(Pommier et al., 2007; Fuhrman et al., 2008). Thisvariability is often attributed to a combination ofenvironmental factors that influence the rate ofgrowth of individual taxa and physical parametersthat prevent different communities from interacting(Crump et al., 2004; Fuhrman et al., 2006, 2008;Lozupone and Knight, 2007; Nemergut et al., 2011).

Of these factors, salinity, temperature and depthappear to be the most important in distinguishingaquatic communities over large spatial scales, inpart because many environmental factors varywith salinity, temperature and depth (for example,light, nutrients, pressure), which results in separa-tion of water masses and thereby communities(Morris et al., 2005; Fuhrman et al., 2008; Carlsonet al., 2009; Treusch et al., 2009; Fortunato andCrump, 2011). On a global scale, Lozupone andKnight (2007) showed that the primary determinantof aquatic microbial community composition wassalinity, whereas Fuhrman et al. (2008) found thatchanges in diversity of marine bacteria acrossa latitudinal gradient were highly correlated totemperature.

Temporal variability in marine and freshwatermicrobial communities is also predictable withinindividual environments. Seasonal shifts in micro-bial community composition have been demon-strated in marine environments, such as theSargasso and Baltic Seas and the English Channel,where succession of microbial communitiescorrelated with changes in mixed layer depth,

Received 26 April 2011; revised 10 August 2011; accepted 19August 2011; published online 20 October 2011

Correspondence: CS Fortunato, Horn Point Laboratory, Universityof Maryland Center for Environmental Science, 2020 Horns PointRd., PO Box 775, Cambridge, MD 21613, USA.E-mail: [email protected]

The ISME Journal (2012) 6, 554–563& 2012 International Society for Microbial Ecology All rights reserved 1751-7362/12

www.nature.com/ismej

temperature and nutrient concentrations throughoutthe year (Morris et al., 2005; Carlson et al., 2009;Gilbert et al., 2009; Andersson et al., 2010). Mixing,temperature and nutrient concentrations are impor-tant factors influencing communities in fresh-water systems as well (Kent et al., 2007; Shadeet al., 2008; Nelson, 2009; Berdjeb et al., 2011).Shade et al. (2008) found distinct communitiesin layers of a stratified lake, where gradients oftemperature, dissolved oxygen and nutrients werepresent. Seasonal succession in both marine andfreshwater has also been shown to be repeatable(Morris et al., 2005; Fuhrman et al., 2006; Carlsonet al., 2009; Crump et al., 2009; Nelson, 2009;Andersson et al., 2010). Crump et al. (2009) showedsynchronous shifts in communities of six arcticrivers strongly correlated with seasonal changesin the environment, suggesting microbial commu-nities may shift in predictable patterns from seasonto season.

Microbial communities are highly diverse, butthe extent and the variability of this diversity infreshwater and marine systems is uncertain. Highthroughput pyrosequencing of PCR-amplified 16SrRNA genes is beginning to resolve the deepdiversity of these systems. Because of the largenumber of sequences per run (B1 million reads),16S amplicon pyrosequencing provides betterresolution of microbial biogeographical patterns,because the depth of diversity captured with eachsample is greater when compared with classicalcommunity fingerprinting techniques (for example,DGGE, T-RFLP, ARISA), which only capture themost dominant species in an environment (Soginet al., 2006). Recent studies have used 16S ampliconpyrosequencing to determine the microbial diversityof many different environments including deep sea,arctic, soil and estuarine communities (Sogin et al.,2006; Galand et al., 2009; Gilbert et al., 2009; Lauberet al., 2009; Andersson et al., 2010).

Microbial community composition and diversityhave been characterized spatially and temporally invarious environments, but rarely have they beenassessed over both spatial and seasonal scales. Using16S amplicon pyrosequencing, we characterizedbacterioplankton communities from 300 water sam-ples collected across the Columbia River coastalmargin over an annual cycle. The coastal waters ofthe Pacific Northwest are highly productive becauseof nutrient delivery from seasonal upwelling andfrom the Juan de Fuca strait and Columbia River(Hickey and Banas, 2003). The biological andphysical processes of these coastal waters arecomplex because of variable winds, remote windforcing, shelf width and submarine canyons (Hickeyand Banas, 2003, 2008; Hickey et al., 2010), whichin turn may differentially affect the composition ofbacterioplankton communities along the Oregon andWashington coasts (Fortunato and Crump, 2011).The Columbia River is the second largest river in theUnited States with a mean annual discharge of

7300 m3 s�1 (Hickey et al., 1998). This significantrelease of freshwater has a strong impact on thechemical, physical and biological characteristics ofthe coastal ocean, including primary and secondaryproduction within the river plume, and differen-tially along the Oregon and Washington coasts(Hickey et al., 2010).

In a previous study in August 2007, the commu-nity fingerprinting technique DGGE was used tobroadly characterize the spatial variation of micro-bial communities in the Columbia River coastalmargin (Fortunato and Crump, 2011). Here we used16S amplicon pyrosequencing to expand on thisearlier dataset by increasing the sample size fourfoldand characterizing communities across multipleseasons using a more resolved spatial scale fromthe river to the deep ocean. We hypothesized thatbecause of the large spatial scale of this study,bacterioplankton communities would separate fromriver to ocean, across salinity, depth and otherenvironmental gradients that vary from fresh tomarine waters. Our results indicate that spatialvariability overwhelmed seasonal trends across theentire sample set, and temporal variability couldonly be resolved within single environment types.

Methods

Water samples were collected from the Oregon andWashington coasts, and the Columbia River andestuary (latitude 44.652 and 47.917, longitude�123.874 and �125.929) as part of the NSF-fundedScience and Technology Center for Coastal MarginObservation and Prediction. Samples were collectedbetween 2007 and 2008 on eight cruises aboardthe R/V Wecoma and R/V Barnes. Aboard the R/VWecoma, water samples were collected from theColumbia River, estuary, plume and two coastalocean lines (Columbia River line, Newport Hydro-line) in August and November of 2007 and April,June, July and September of 2008 (Figure 1). Forcoastal lines, samples were taken at three depths perstation (surface, within thermocline and bottom).Plume samples were taken at two depths (surfaceand bottom) in 2007 and four depths (surface, belowplume, within thermocline and bottom) in 2008.In the estuary, samples were collected based onthe location of the salt gradient in both the north andsouth channels of the river. Samples were collectedacross the salt gradient from 0 to 30. Samples werecollected using a conductivity-temperature-depth(CTD) rosette water sampler with 10-liter Niskinbottles. With each CTD cast, depth profiles ofsalinity, temperature (1C), turbidity (NTU), oxygen(mgl�1) and chlorophyll fluorescence were recorded.Water samples aboard the R/V Barnes were collectedusing a high-volume low-pressure pump oversalinity gradients in the estuary in August 2007and July 2008. For all samples, surface was definedbetween 1 and 2 m depth, and bottom was defined

Spatial overwhelms seasonal community patternsCS Fortunato et al

555

The ISME Journal

between 1–5 m above sediment. Data from CTDfluorescence and temperature sensors were usedto determine exact sampling depths for watercollected at the chlorophyll maximum and withinthe thermocline.

DNA samples (1–6 l per sample) were collected,preserved and extracted as described previously(Fortunato and Crump, 2011) using methodsadapted from Zhou et al. (1996) and Crump et al.(2003). Extracted DNA was PCR-amplified usingprimers targeting bacterial 16S ribosomal RNAgenes. Each sample was assigned a uniquelybarcoded reverse primer and amplified in fourreplicate 20-ml reactions (Hamady et al., 2008).Primers used for amplification were bacteria-specificprimers focusing on the V2 region, 27F with 454BFLX linker (50-GCCTTGCCAGCCCGCTCAGTCAGRGTTTGATYMTGGCTCAG-30) and 338R with 454Alinker and a unique 8-bp barcode, denoted by N inprimer sequence (50-GCCTCCCTCGCGCCATCAGNNNNNNNCATGCWGCCWCCCGTAGGWGT-30) (modi-fied from Hamady et al., 2008). Replicate amplifica-tions were combined, purified and normalized usingInvitrogen SequelPrep normalization plates (Invitro-gen, Carlsbad, CA, USA). In all, 5 ml from eachsample was combined into a single tube and sent forpyrosequencing on a Roche-454 FLX pyrosequencerat Engencore at the University of South Carolina(http://engencore.sc.edu/).

Sequence data were processed using two differentmethods: (1) Manual global alignment and removalof pyrosequencing errors using ARB (Ludwiget al., 2004) and MOTHUR software (Schloss et al.,2009), and (2) Denoising and pairwise alignment

using the QIIME (v.1.2.0) software package(Caporaso et al., 2010).

For the first method, raw sequences were sortedand quality controlled (minimum length 150 bp, noambiguous bases) using the Ribosomal DatabaseProject Pyro tools (Cole et al., 2005). A referencesequence database was created using the communityanalysis program MOTHUR (Schloss et al., 2009)consisting of unique sequences from the overalldataset. These unique sequences were imported intoARB and manually aligned. Extra bases commonlyadded in pyrosequencing (that is, pyronoise) wereplaced in gaps added to the alignment. Once themanual alignment was completed, sequences weretrimmed to E. coli basepair positions 136–335 andwere exported using a 3% basepair frequency filterto mask insertions, but include variable bases.This reference dataset of manually aligned uniquesequences was then used to align the entire datasetusing MOTHUR. Our approach removed insertionsfrom pyrosequencing, but did not repair deletionsof bases, which were included in downstreamanalyses. Operational taxonomic units (OTUs) weredetermined based on 97% sequence similarity usingMOTHUR.

For the second method using QIIME, sequenceswere quality controlled using the Split_Libraries.pyscript with default settings (minimum length 200,maximum length 1000, minimum mean qualityscore 25, maximum ambiguous bases 0, maximumhomopolymer length 6, maximum primer mismatch0). To account for pyronoise, the remainingsequences were denoised using the denoiser.pyscript with the ‘fast’ method and default settings.Sequences were then clustered using the pick_otus.py script with the uclust method (97% sequ-ence similarity). Potentially chimeric sequenceswere identified among representative sequencesfrom each OTU with ChimeraSlayer, and a total of3952 sequences composing 196 OTUs were elimi-nated from the dataset.

For both methods, relative abundance was calcu-lated for the OTUs in each sample and used tocalculate pairwise similarities among samples usingthe Bray–Curtis similarity coefficient (Legendre andLegendre, 1998). We also calculated pairwise simi-larities among samples using both weighted andunweighted UNIFRAC metrics (Lozupone et al.,2006), but the results were nearly identical to thosebased on Bray–Curtis, and so are not presented.Bray–Curtis similarity matrices were visualizedusing multiple dimensional scaling (MDS) dia-grams, a form of ordination. Analysis of SimilarityStatistics (ANOSIM) was calculated to test thesignificance of differences among a priori sampl-ing groups based on environmental parameters.Similarity matrices, MDS diagrams and ANOSIMstatistics were carried out using PRIMER v6 forWindows (PRIMER-E Ltd, Plymouth, UK).

Alpha diversity for samples was calculatedusing MOTHUR. The number of sequences was

Figure 1 Map of the Oregon and Washington coast. Inset depictsColumbia River estuary and plume region. Dotted line denotesapproximate location of shelf break.

Spatial overwhelms seasonal community patternsCS Fortunato et al

556

The ISME Journal

normalized before calculation by randomly selectingthe same number of sequences per sample, basedupon the sample with the least number of sequences(n¼ 209 sequences). The taxonomy of OTUs identi-fied was determined using the Ribosomal DatabaseProject Classifier tool. Taxonomic assignments withless than 80% confidence were marked as unknown.A total of 306 samples were analyzed overall. Thisnumber was reduced to 300 as samples with a lownumber of sequences were removed.

All sequences can be downloaded from the NCBISequence Read Archive database under the acces-sion number SRP006412. In addition, a Supplemen-tary Table containing sample metadata conformingto MIMARKS standards has also been provided(Supplementary Table S1).

Results

Comparison of the two sequence analysis methodsshowed that the overall patterns of microbialcommunity structure for this study are highlyrobust, as both spatial and temporal patterns inbeta-diversity were the same for both methods. Thenumber of OTUs identified by the QIIME analysis(8039) was slightly lower than that of the ARB/MOTHUR analysis (9389), but this was becausefewer sequences passed the initial QIIME qualitycontrol step due to different quality control para-meters, including maximum homopolymer lengthand primer mismatches. Because the patterns ofcommunity variability were comparable, the resultspresented are based on the QIIME sequence analysisprotocol.

Bacterioplankton communities separated intoseven distinct groups (ANOSIM, Po0.001): river,estuary, plume, epipelagic, mesopelagic, shelfbottom and slope bottom. The plume groupconsisted of coastal surface samples with salinityless than 31, the epipelagic group included coastalsurface and chlorophyll maximum samples (averagedepth¼ 8 m), the mesopelagic group consisted ofcoastal samples within and below the thermocline(average depth¼ 44 m), the shelf bottom groupconsisted of bottom samples with depth less than350 m and the slope bottom group consisted of

bottom samples deeper than 850 m. Percent similar-ity for all samples was 22.9% (±15.3%) with a rangefrom 0% to 74.8% similarity. Similarity values werehigher within groups than between groups (Table 1).

A MDS diagram of all 300 samples based on Bray–Curtis similarity values (Figure 2) depicts the sevengroups based on location in the system. Groupsseparate along two axes that form a V-shapedarrangement of microbial communities. The firstaxis is clearly related to salinity and the second isrelated to depth. A strong correlation was shownbetween Dimension 1 and salinity, with a Spear-man’s rho value of �0.83 (Po0.001, Figure 3). Aweaker relationship was observed between Dimen-sion 2 and sample depth (r¼�0.62, Po0.001 forDimension 2 axis and depth), although this relation-ship improved when river and estuary samples wereomitted (r¼�0.76, Po0.001).

Spatial variation in communities based on sam-pling location is readily apparent in Figure 2.Temporal variation, however, appears to be over-whelmed by the strong spatial gradients of salinityand depth. Temporal variation was only detectablewhen each spatial group was analyzed separately.For river, estuary and plume samples, a seasonaltrend is apparent from river to ocean (Figure 4). In

Table 1 Percent similarity values within and between groups ±standard deviation (ANOSIM: Po0.001) as determined by Bray–Curtissimilarity coefficient

River Estuary Plume Epipelagic Mesopelagic Shelf bottom Slope bottom

River 33.6±11.2Estuary 17.8±11.7 24.9±11.1Plume 2.6±4.1 19.2±11.4 36.5±10.5Epipelagic 0.1±0.2 15.4±11.2 35.0±9.2 38.4±9.9Mesopelagic 0.4±1.5 13.1±10.0 27.3±11.1 33.4±11.8 37.2±11.5Shelf bottom 0.2±0.3 10.3±8.6 17.9±9.5 22.8±10.9 32.2±12.1 41.8±12.6Slope bottom 0.2±0.3 4.3±4.0 6.2±5.8 8.7±6.3 16.4±9.6 24.7±9.0 50.3±6.6

Abbreviation: ANOSIM, Analysis of Similarity Statistics.

Figure 2 Multiple dimensional scaling diagram of percentsimilarities for all 300 samples. Bacterioplankton communitieswere separated into seven groups based on location across salinityand depth gradients (ANOSIM: Po0.001, Stress: 0.12).

Spatial overwhelms seasonal community patternsCS Fortunato et al

557

The ISME Journal

the river, three communities are visible: spring,freshet-early summer and late summer-fall. In theestuary, seasonal clustering of communities was notas clear, although communities did split into twosignificant clusters (ANOSIM, Po0.001), an early-year community, encompassing samples from Aprilto July, and a late-year community, encompassingsamples from August to November. These sametwo communities, early and late, are also presentin the plume (ANOSIM, Po0.001). The seasonalpattern in the other groups is less discernable.There was significant seasonal variation in theshelf bottom and epipelagic groups according tothe ANOSIM statistics, but these patterns couldnot be discerned in the individual MDS diagramsbecause of the large amount of variability withineach group. There was no significant temporalpattern in the slope bottom or mesopelagic groups.

Most sequences in the dataset were related to thephyla Proteobacteria (44.7%) and Bacteriodetes(33.6%). Within the Proteobacteria, Alpha (21.2%),Gamma (17.0%), Beta (2.6%) and Delta (0.4%)were present. In the Bacteriodetes, Flavobacteriawas the largest group, with 55 915 sequences makingup 28% of the total dataset. The most abundantOTU belonged to the SAR-11 clade and consistedof 16 635 sequences. Overall, SAR-11 made up11.3% of the dataset with a total of 22 454 sequencesbelonging to 208 OTUs. The second largest OTUwas a Gammaproteobacteria with 13 137 sequences.Cyanobacteria was a small percentage of the totaldataset, only 1.8%, but constituted as much as19% of sequences in epipelagic samples collectedoff the shelf. More specific taxonomic informa-tion for each of the seven spatial groups can befound in the Supplementary Material (Table S2,Figure S1).

To better understand community composition, weclassified each of the 8039 OTUs in this study

based on the location in the system where theyexhibited their maximum average relative abun-dance in pooled sequences (Figure 5). For example,if OTU-1 was most abundant in the plume (basedon its relative abundance within each pool ofsequences from the seven groups), it was classifiedas a plume OTU. Results suggest mixing of watermasses and microbial communities from estuary tothe shelf bottom. The river and slope bottom groupsappear to be end members in the system, as most ofthe river and slope bottom sequences are found

Figure 3 Correlation of Dimension 1 for the 300 samples fromFigure 2 and salinity. A Spearman’s rho value of �0.83 (Po0.001)indicates a strong relationship between salinity and bacterialcommunity variation.

Figure 4 Seasonal multiple dimensional scaling diagram ofriver, estuary and plume. River displays three seasonal commu-nites, which cluster into two communities, early (April–July) andlate (August–November), in the estuary and plume. Stress¼0.04,0.15 and 0.17 for river, estuary and plume, respectively.

Spatial overwhelms seasonal community patternsCS Fortunato et al

558

The ISME Journal

only in their respective locations. The estuarycommunity is primarily a mix of sequences belong-ing to river and estuarine OTUs, with some addi-tion from the plume and epipelagic. In the plume,however, plume sequences are mostly classifiedas being from plume, epipelagic and mesopelagicOTUs with few sequences coming from river orestuary OTUs.

We mapped the relative abundance of the topOTU from each of the seven spatial groups (based onaverage relative abundance per group) using theordination of Figure 2. These bubble plots show thatthe top OTUs for each group are most abundant insamples from their location and less abundant inneighboring locations (Supplementary Figure S2).The top OTUs for the estuary and the river displaysome seasonality, with the largest abundancesoccurring in only one or two seasons (for example,June and July 2008 for the estuary).

Alpha-diversity varied across the spatial groups(Figure 6). The river and slope bottom groups hadthe highest and third highest average diversity(Chao1¼ 1104 and 868, respectively), indicatingthe presence of many more endemic taxa withinthese two environments, and showing further thatfreshwater and deep ocean represent end membersin this study. As water mixes from the river tothe coastal surface ocean, diversity measurementsdecrease to the lowest diversity in the epipelagicgroup (Chao1¼ 380). Diversity then increased fromsurface to the deep ocean, with the mesopelagic,shelf bottom, slope bottom groups each havinga higher diversity than the previous. Diversitymeasurements show that when water mixes fromfresh to salt and from deep to surface, taxa arereduced in abundance beyond our limit of detectionand thus community composition becomes morestreamlined in the coastal surface.

Discussion

Previous studies of variability and diversity inbacterioplankton communities are restricted tosingle dimensions, focusing on long-term timeseries, depth profiles or horizontal surveys acrossenvironmental gradients (Morris et al., 2005;Hewson et al., 2006; Lozupone and Knight, 2007;Pommier et al., 2007; Fuhrman et al., 2008; Gilbertet al., 2009; Treusch et al., 2009; Andersson et al.,2010; Nemergut et al., 2011). Here we present adataset that compares bacterioplankton communitycomposition in all three of these dimensions:spatially from river to surface ocean, by depth fromsurface to deep ocean, and through time seasonallyover an annual cycle. This large-scale biogeographi-cal analysis was enabled by the use of 16S ampliconpyrosequencing, which assesses diversity throughDNA sequencing of hundreds of thousands of PCR-amplified gene copies. Previous 16S ampliconpyrosequencing studies focused on deep samplingof small numbers of samples, allowing for character-ization of the ‘rare biosphere’ but only at limitedspatial and temporal scales (Galand et al., 2009;Gilbert et al., 2009; Andersson et al., 2010; Kirch-man et al., 2010). In this study, we took a differentapproach to characterizing bacterioplankton com-munities by applying 16S amplicon pyrosequencingto ten times the number of samples seen in previousstudies. Sequencing more samples produces fewersequences per sample and limits the resolution ofthe rare biosphere. However, the greater number ofsamples in this study (n¼ 300) led to the discoveryof robust spatial patterns from river to ocean andseasonal shifts that may not have been observed iffewer samples were sequenced. Based on a previouscommunity fingerprinting study of 71 samples fromAugust 2007 using DGGE, we found that commu-

Figure 5 Percentage of sequences in OTUs classified by location.Slope bottom and river groups represent end members in thesystem. Rare category represents sequences belonging to OTU thatmake up to less than 0.1% of the total number of sequences fromeach corresponding location.

Figure 6 Average Chao1 index per group ± standard deviationas determined using MOTHUR (v.1.15.0). OTU number wasnormalized to the sample with the smallest number of sequences(n¼ 209 sequences).

Spatial overwhelms seasonal community patternsCS Fortunato et al

559

The ISME Journal

nities separated into just five groups defined bylocation across salinity and depth gradients (For-tunato and Crump, 2011). With the addition of overfour times the number of samples, in this study wewere able to further resolve the spatial patterns ofbacterioplankton communities into seven distinctgroups across steep salinity and depth gradients inaddition to determining temporal variability.

Salinity and depth changed significantly from theColumbia River to the deep ocean, and these factorsappear to strongly influence the composition ofbacterioplankton communities. In contrast, tempor-al variability in bacterioplankton communities wasrelatively small, and was obscured by the spatialvariability in communities across environments inthe coastal zone. Several studies of coastal zonebacterioplankton identify time as the principle axisof community variability (Stepanauskas et al., 2003;Fuhrman et al., 2006; Kan et al., 2006; Gilbert et al.,2009), but these studies were restricted to oneenvironmental type (for example, estuaries or afixed coastal station) within which spatial varia-bility of bacterioplankton communities was limited.Few studies address temporal variability acrossmany different habitats, so it was difficult tocompare our results with other studies. However,one study by Kirchman et al. (2010) identified asimilar pattern among 11 surface water samples inwhich winter/summer differences in Arctic Oceanbacterioplankton communities was minimal com-pared with spatial variability across their samplingrange. Thus, although temporal variability mayoccur within many marine habitats, it is clear thatstructuring environmental factors (for example,salinity, depth) dominate over seasonal changes indetermining community composition.

Spatial differentiation among samples was highlycorrelated with salinity, confirming the observationsof two global meta-analyses of microbial diversitybased on 16S rRNA gene sequences (Lozupone andKnight, 2007; Tamames et al., 2010). In one of thesestudies, Lozupone and Knight (2007) found thatsalinity was the primary environmental determinantfor community composition across marine, fresh-water, sediment and soil environments, more sothan temperature, pH or other environmentalfactors. In the coastal marine environment, salinitycontributes to density gradients that physicallyseparate water masses and their resident microbialcommunities. However, the degree to which thesewater masses are separated depends on the magni-tude of mixing by river flow, tides, upwelling,surface winds, and so on. This mixing from freshto marine or from surface to deep leads to theformation of communities in mixing zones thatcomprise bacterioplankton populations frommultiple water masses. For example in the ColumbiaRiver estuary, the flushing rate exceeds the doublingtime of bacterioplankton populations, thus adistinct free-living estuarine community is unableto form (Crump et al., 1999). Our study confirmed

this observation, demonstrating that estuarinebacterioplankton communities are composed ofpopulations from the river and the coastal ocean(Figure 5). We also identified significant overlap incommunities across environmental gradients in thecoastal ocean including the plume, epipelagic,mesopelagic and shelf bottom environments (Figure5 and Supplementary Figure S2), although it isunclear whether this is the result of mixing or thepresence of generalist organisms that thrive indifferent environments.

Coastal bacterioplankton communities correlatedwith depth from the surface to the deep ocean,despite the fact that samples were collected overmultiple seasons and at sampling sites as much as150 km apart. Salinity varies with depth, as do manyother environmental parameters including tempera-ture, light and nutrients. We therefore are treatingdepth here as a proxy for many factors that vary inthe vertical dimension. The vertical structuring ofbacterioplankton communities in the ocean hasbeen demonstrated in many studies and has beenlinked to changes in hydrostatic pressure as well aswater mass properties (Lee and Fuhrman, 1991;Morris et al., 2005; Blumel et al., 2007; Carlsonet al., 2009; Treusch et al., 2009). For example,Treusch et al. (2009) found that Sargasso Seabacterial communities separated into surface (upper40 m), deep chlorophyll maximum and upper meso-pelagic communities. We also observed a separationof the epipelagic and upper mesopelagic commu-nities, but not between surface and chlorophyllmaximum samples, possibly because the mixedlayer depth (5–56 m) was, in general, shallower thanthat of the Sargasso Sea (o50–350 m)(Carlson et al.,2009; Treusch et al., 2009). Treusch et al. (2009)attributed separation of these communities to stra-tification and seasonal mixing in the upper watercolumn. The coastal zone of the Pacific North-west experiences seasonal upwelling, and thus amixing of communities from bottom to surface. Thedegree of mixing is evident in Figure 5, where themesopelagic group is actually a mix of populationsfrom the bottom and surface. In July 2008 duringstrong upwelling, near-shore surface samples fromthe Newport Hydroline contained a higher propor-tion (23%) of sequences belonging to shelf bottomand slope bottom OTUs than during other times in2008 (5%). Also during that month, the mostabundant estuary-classified OTU was found in someshelf bottom samples (Supplementary Figure S2),indicating a possible exchange between these twoenvironments.

Temporal variability could only be resolvedwithin some environments. Seasonal changes wereobserved in the river, estuary and plume environ-ments. In the river, there were three separate groups,spring, freshet-early summer and late summer-fall,corresponding to seasonal changes in ColumbiaRiver discharge, where maximum discharge occursin late spring and is minimum in late summer to

Spatial overwhelms seasonal community patternsCS Fortunato et al

560

The ISME Journal

early fall (Prahl et al., 1998). In the plume, seasonalupwelling strongly influences temperature andnutrient concentrations, and thereby production inthe plume (Hickey et al., 2010). Thus, plumecommunity composition is tightly linked to thephysical processes occurring along the coast. Theseasonality of the estuary community then can beattributed to a combination of both river and coastalprocesses. The periods of maximum and minimumdischarge of the river correspond to the two seasonalbacterioplankton groups seen in the estuary, early(April–July) and late (August–November). Duringtimes of high river flow, the estuarine community isshaped by the river and when river flow is at aminimum, community composition is influencedmore by the plume and coastal ocean.

River and deep ocean (slope bottom group) appearto be end members in this system in that theycontribute populations to nearby environments, butreceive little to no contributions themselves(Figure 5). In the other five groups there was tremen-dous overlap in community composition fromestuary to shelf bottom, suggesting dynamic ex-change of communities through advection andmixing. Within each group there also appeared tobe environment-specific communities, based onmaximum relative abundance (Figure 5). In theplume, 37% of plume sequences were classified asbelonging to plume OTUs, indicating the presenceof a plume-specific community. Additionally, only5% of plume sequences were from the river andestuary, whereas 36% came from epipelagic andmesopelagic OTUs, indicating the plume commu-nity is comprised more of coastal populations thanbacteria flushed from the estuary. As mentionedpreviously, the plume is highly productive becauseof nutrient delivery from the river and coastalupwelling (Hickey et al., 2010) and as primaryproduction increases in the plume, different epipe-lagic taxa could increase depending on availabilityand quality of organic matter. This would result ina different combination of bacterial populationsand a clear distinction between the plume andepipelagic communities. We speculate then thateach spatial group, from estuary to shelf bottom,contains bacterioplankton populations that arebroadly distributed across environments, but eachgroup supports a different combination of thesebacteria, creating distinct communities within eachenvironment.

16S amplicon pyrosequencing, like any othermolecular technique, is prone to errors and it isimportant to analyze sequences in a way thataccurately assesses community patterns. Analyzing16S amplicon pyrosequencing data is difficultbecause of sequencing errors termed ‘pyronoise’,which may artificially increase the number of OTUsobserved. In Kunin et al. (2009), the authors PCR-amplified a 300 bp region of the 16S rRNA gene froma known cultured E. coli strain and then pyrose-quenced it. The results returned a largely inflated

number of OTUs, showing that pyrosequencingerrors may lead to a gross overestimation of thenumber of OTUs in a sample. An increase in thenumber of OTUs leads to inflated alpha diversitywithin samples, and greater beta diversity betweensamples. We found that global alignment combinedwith manual removal of pyronoise insertions wascomparable in total OTU number, alpha-diversityand beta-diversity patterns to analysis using aQIIME analysis pipeline that includes denoising(denoiser.py) and pairwise sequence alignment(uclust). We also found that removing the pyronoiseis crucial for minimizing the total number of OTUsand overall sequencing errors. To demonstrate this,we globally aligned our sequences using a referencedatabase from SILVA (Pruesse et al., 2007) andfound that although our beta-diversity patterns werecomparable, the OTU number and alpha diversityestimates were nearly twice than that of ourprevious methods (data not shown). It is importantthen that pyrosequencing datasets be subjected torigorous quality checking and denoising, in order toaccurately assess both the overall community pat-terns and the rare biosphere.

Acknowledgements

This study was carried out within the context of theScience and Technology Center for Coastal MarginObservation & Prediction supported by the NationalScience Foundation (grant number OCE-0424602 toAntonio Baptista). We would like to thank the captainsand crews of the R/V Wecoma and R/V Barnes, chiefscientists Murray Levine, Tawnya Peterson and JoeNeedoba, as well as Suzanne DeLorenzo, Peter Kahn andDaniel Murphy for their help in sample collection. Wealso thank Joe Jones and John Busch for 454-pyrosequen-cing support, and Patrick Schloss and Jack Gilbert for theirhelp in pyrosequence analysis.

References

Andersson AF, Riemann L, Bertilsson S. (2010). Pyro-sequencing reveals contrasting seasonal dynamics oftaxa within Baltic Sea bacterioplankton communities.ISME J 4: 171–181.

Berdjeb L, Ghiglione JF, Jacquet S. (2011). Bottom-upversus top-down control of hypo- and epilimnion free-living bacterial community structures in two neigh-boring freshwater lakes. Appl Environ Microbiol 77:3591–3599.

Blumel M, Suling J, Imhoff JF. (2007). Depth-specificdistribution of Bacteroidetes in the oligotrophiceastern Mediterranean Sea. Aquat Microb Ecol 46:209–224.

Caporaso JG, Kuczynski J, Stombaugh J, Bittinger K,Bushman FD, Costello EK et al. (2010). QIIME allowsanalysis of high-throughput community sequencingdata. Nat Methods 7: 335–336.

Carlson CA, Morris R, Parsons R, Treusch AH, Giovannoni SJ,Vergin K. (2009). Seasonal dynamics of SAR11

Spatial overwhelms seasonal community patternsCS Fortunato et al

561

The ISME Journal

populations in the euphotic and mesopelagic zones ofthe northwestern Sargasso Sea. ISME J 3: 283–295.

Cole JR, Chai B, Farris RJ, Wang Q, Kulam SA,McGarrell DM et al. (2005). The RibosomalDatabase Project (RDP-II): sequences and tools forhigh-throughput rRNA analysis. Nucleic Acids Res 33:D294–D296.

Crump BC, Armbrust EV, Baross JA. (1999). Phylogeneticanalysis of particle-attached and free-living bacterialcommunities in the Columbia River, its estuary, andthe adjacent coastal ocean. Appl Environ Microbiol 65:3192–3204.

Crump BC, Hopkinson CS, Sogin ML, Hobbie JE. (2004).Microbial biogeography along an estuarine salinitygradient: combined influences of bacterial growthand residence time. Appl Environ Microbiol 70:1494–1505.

Crump BC, Kling GW, Bahr M, Hobbie JE. (2003).Bacterioplankton community shifts in an arctic lakecorrelate with seasonal changes in organic mattersource. Appl Environ Microbiol 69: 2253–2268.

Crump BC, Peterson BJ, Raymond PA, Amon RMW,Rinehart A, McClelland JW et al. (2009). Circumpolarsynchrony in big river bacterioplankton. Proc NatlAcad Sci USA 106: 21208–21212.

Fortunato CS, Crump BC. (2011). Bacterioplankton com-munity variation across river to ocean environmentalgradients. Microb Ecol 62: 374–382.

Fuhrman JA, Hewson I, Schwalbach MS, Steele JA, BrownMV, Naeem S. (2006). Annually reoccurring bacterialcommunities are predictable from ocean conditions.Proc Natl Acad Sci USA 103: 13104–13109.

Fuhrman JA, Steele JA, Hewson I, Schwalbach MS,Brown MV, Green JL et al. (2008). A latitudinaldiversity gradient in planktonic marine bacteria. ProcNatl Acad Sci USA 105: 7774–7778.

Galand PE, Casamayor EO, Kirchman DL, Lovejoy C.(2009). Ecology of the rare microbial biosphereof the Arctic Ocean. Proc Natl Acad Sci USA 106:22427–22432.

Gilbert JA, Field D, Swift P, Newbold L, Oliver A, Smyth Tet al. (2009). The seasonal structure of microbialcommunities in the Western English Channel. EnvironMicrobiol 11: 3132–3139.

Hamady M, Walker JJ, Harris JK, Gold NJ, Knight R. (2008).Error-correcting barcoded primers for pyrosequencinghundreds of samples in multiplex. Nat Methods 5:235–237.

Hewson I, Steele JA, Capone DG, Fuhrman JA. (2006).Temporal and spatial scales of variation in bacterio-plankton assemblages of oligotrophic surface waters.Mar Ecol Prog Ser 311: 67–77.

Hickey BM, Banas NS. (2003). Oceanography of the USPacific Northwest Coastal Ocean and estuarieswith application to coastal ecology. Estuaries 26:1010–1031.

Hickey BM, Banas NS. (2008). Why is the Northern end ofthe California current system so productive? Oceano-graphy 21: 90–107.

Hickey BM, Kudela RM, Nash JD, Bruland KW,Peterson WT, MacCready P et al. (2010). River influ-ences on shelf ecosystems: introduction and synthesis.J Geophys Res-Oceans 115: 1–26.

Hickey BM, Pietrafesa LJ, Jay DA, Boicourt WC. (1998).The Columbia River plume study: subtidal variabilityin the velocity and salinity fields. J Geophys ResOceans 103: 10339–10368.

Kan JJ, Crump BC, Wang K, Chen F. (2006). Bacterio-plankton community in Chesapeake Bay: predic-table or random assemblages. Limnol Oceanogr 51:2157–2169.

Kent AD, Yannarell AC, Rusak JA, Triplett EW,McMahon KD. (2007). Synchrony in aquatic microbialcommunity dynamics. ISME J 1: 38–47.

Kirchman DL, Cottrell MT, Lovejoy C. (2010). Thestructure of bacterial communities in the westernArctic Ocean as revealed by pyrosequencing of 16SrRNA genes. Environ Microbiol 12: 1132–1143.

Kunin V, Engelbrektson A, Ochman H, Hugenholtz P.(2009). Wrinkles in the rare biosphere: pyrosequen-cing errors can lead to artificial inflation of diversityestimates. Environ Microbiol 12: 118–123.

Lauber CL, Hamady M, Knight R, Fierer N. (2009).Pyrosequencing-based assessment of soil pH as apredictor of soil bacterial community structure atthe continental scale. Appl Environ Microbiol 75:5111–5120.

Lee SH, Fuhrman JA. (1991). Spatial and temporalvariation of natural bacterioplankton assemblagesstudied by total genomic DNA cross-hybridization.Limnol Oceanogr 36: 1277–1287.

Legendre P, Legendre L. (1998). Numerical Ecology. Else-vier Science B.V.: Amsterdam.

Long RA, Azam F. (2001). Microscale patchiness ofbacterioplankton assemblage richness in seawater.Aquat Microb Ecol 26: 103–113.

Lozupone C, Hamady M, Knight R. (2006). UniFrac—anonline tool for comparing microbial communitydiversity in a phylogenetic context. BMC Bioinform7: 371.

Lozupone CA, Knight R. (2007). Global patterns inbacterial diversity. Proc Natl Acad Sci USA 104:11436–11440.

Ludwig W, Strunk O, Westram R, Richter L, Meier H,Yadhukumar et al. (2004). ARB: a software environmentfor sequence data. Nucleic Acids Res 32: 1363–1371.

Morris RM, Vergin KL, Cho JC, Rappe MS, Carlson CA,Giovannoni SJ. (2005). Temporal and spatial responseof bacterioplankton lineages to annual convectiveoverturn at the Bermuda Atlantic Time-series Studysite. Limnol Oceanogr 50: 1687–1696.

Nelson CE. (2009). Phenology of high-elevation pelagicbacteria: the roles of meteorologic variability, catch-ment inputs and thermal stratification in structuringcommunities. ISME J 3: 13–30.

Nemergut DR, Costello EK, Hamady M, Lozupone C,Jiang L, Schmidt SK et al. (2011). Global patterns inthe biogeography of bacterial taxa. Environ Microbiol13: 135–144.

Pommier T, Canback B, Riemann L, Bostrom KH, Simu K,Lundberg P et al. (2007). Global patterns of diversityand community structure in marine bacterioplankton.Mol Ecol 16: 867–880.

Prahl FG, Small LF, Sullivan BA, Cordell J, Simenstad CA,Crump BC et al. (1998). Biogeochemical gradients inthe lower Columbia River. Hydrobiologia 361: 37–52.

Pruesse E, Quast C, Knittel K, Fuchs BM, Ludwig WG,Peplies J et al. (2007). SILVA: a comprehensive onlineresource for quality checked and aligned ribosomalRNA sequence data compatible with ARB. NucleicAcids Res 35: 7188–7196.

Schloss PD, Westcott SL, Ryabin T, Hall JR, Hartmann M,Hollister EB et al. (2009). Introducing mothur: open-source, platform-independent, community-supported

Spatial overwhelms seasonal community patternsCS Fortunato et al

562

The ISME Journal

software for describing and comparing microbial com-munities. Appl Environ Microbiol 75: 7537–7541.

Shade A, Jones SE, McMahon KD. (2008). The influence ofhabitat heterogeneity on freshwater bacterial commu-nity composition and dynamics. Environ Microbiol 10:1057–1067.

Sogin ML, Morrison HG, Huber JA, Mark Welch D,Huse SM, Neal PR et al. (2006). Microbial diversityin the deep sea and the underexplored ‘rare bio-sphere’. Proc Natl Acad Sci USA 103: 12115–12120.

Stepanauskas R, Moran MA, Bergamaschi BA,Hollibaugh JT. (2003). Covariance of bacterioplankton

composition and environmental variables in a tempe-rate delta system. Aquat Microb Ecol 31: 85–98.

Tamames J, Abellan JJ, Pignatelli M, Camacho A, Moya A.(2010). Environmental distribution of prokaryotictaxa. BMC Microbiol 10: 85.

Treusch AH, Vergin KL, Finlay LA, Donatz MG,Burton RM, Carlson CA et al. (2009). Seasonality andvertical structure of microbial communities in anocean gyre. ISME J 3: 1148–1163.

Zhou JZ, Bruns MA, Tiedje JM. (1996). DNA recovery fromsoils of diverse composition. Appl Environ Microbiol62: 316–322.

Supplementary Information accompanies the paper on The ISME Journal website (http://www.nature.com/ismej)

Spatial overwhelms seasonal community patternsCS Fortunato et al

563

The ISME Journal

Slope Bottom

Shelf Bottom

Mesopelagic Epipelagic Plume Estuary River

VerrucomicrobiaPlanctomycetesDeltaproteobacteriaCyanobacteriaBetaproteobacteriaActinobacteriaGammaproteobacteriaBacteroidetesAlphaproteobacteriaOtherUnknown

0

10

20

30

40

50

60

70

80

90

100

Perc

enta

ge o

f se

quen

ces

from

eac

h lo

catio

ncl

assi

fied

by t

axon

omy

usin

g th

e RD

P Cl

assi

fier

tool

Supplemental Figure 1: Percentage of sequences for each spatial group classified using the RDP Classifier toolwith an 80% confidence threshold.

-2

-1.5

-1

-0.5

0

0.5

1

1.5

-1 -0.5 0 0.5 1 1.5 2 2.5 3

Dimension 1

Dim

ensi

on 2

-1 -0.5 0 0.5 1 1.5 2 2.5 3-2

-1.5

-1

-0.5

0

0.5

1

1.5

Dim

ensi

on 2

2.5

-2

-1.5

-1

-0.5

0

0.5

1

1.5

Dim

ensi

on 2

-1 -0.5 0 0.5 1 1.5 2 3

-2

-1.5

-1

-0.5

0

0.5

1

1.5

Dim

ensi

on 2

-1 -0.5 0 0.5 1 1.5 2 3-1.5 2.5

-2

-1.5

-1

-0.5

0

0.5

1

1.5

-1 -0.5 0 0.5 1 1.5 2 32.5

-2

-1.5

-1

-0.5

0

0.5

1

1.5

-1 -0.5 0 0.5 1 1.5 2 32.5

-2

-1.5

-1

-0.5

0

0.5

1

1.5

-1 -0.5 0 0.5 1 1.5 2 32.5

-2

-1.5

-1

-0.5

0

0.5

1

1.5

-1 -0.5 0 0.5 1 1.5 2 32.5

Dimension 1

Unknown Bacteria

All Samples

Gammaproteobacteria

Pelagibacter

Rhodobacteraceae

Flavobacteriaceae

Oceanospirillales

Flavobacterium

Supplemental Figure 2: Spatial distribution of the most abundant OTUs based on average relative abundance in each spatialgroup across the ordination of Figure 2. Taxonomic identification for each OTU is listed to the furthest classification possibleas determined by the RDP Classifier tool with an 80% confidence threshold. Stress = 0.12

0.04

0.16

0.28

0.4

RiverEstuaryPlumeEpipelagicMesopelagicShelf BottomSlope Bottom

Relative Abundance

Spatial Group

Supplemental Table 2: Most abundant OTUs within each spatial group based on average relative abundance. Taxonomy was determined by the RDP Classifier tool with an 80% confidence threshold.

Environment Taxonomy (Phylum, Class, Order, Family, Genus) Average relative abundance within group (%) River Bacteroidetes, Flavobacteria, Flavobacteriales, Flavobacteriaceae, Flavobacterium 4.1 Proteobacteria, Betaproteobacteria, Burkholderiales, Comamonadaceae 4.0 Proteobacteria 3.4 Unknown Bacteria 3.3 Actinobacteria, Actinobacteridae, Actinomycetales 2.7 Proteobacteria, Betaproteobacteria, Methylophilales, Methylophilaceae 2.4 Actinobacteria, Actinobacteridae, Actinomycetales 2.4 Verrucomicrobia 1.8 Bacteroidetes, Flavobacteria, Flavobacteriales, Flavobacteriaceae, Flavobacterium 1.6 Actinobacteria 1.3 Sum: 27.0 Estuary Proteobacteria, Gammaproteobacteria, Oceanospirillales 2.5 Bacteroidetes, Flavobacteria, Flavobacteriales, Flavobacteriaceae 2.4 Proteobacteria, Alphaproteobacteria, Rhodobacterales, Rhodobacteraceae 1.2 Bacteroidetes, Flavobacteria, Flavobacteriales, Cryomorphaceae, Fluviicola 1.2 Bacteroidetes, Flavobacteria, Flavobacteriales, Flavobacteriaceae 1.2 Proteobacteria, Alphaproteobacteria, Rhodobacterales, Rhodobacteraceae 1.1 Proteobacteria, Betaproteobacteria, Burkholderiales, Comamonadaceae 1.1 Bacteroidetes, Flavobacteria, Flavobacteriales, Flavobacteriaceae, Flavobacterium 0.9 Bacteroidetes 0.8 Bacteroidetes, Flavobacteria, Flavobacteriales, Flavobacteriaceae 0.7 Sum: 13.0 Plume Bacteroidetes, Flavobacteria, Flavobacteriales, Flavobacteriaceae 3.9 Bacteroidetes, Flavobacteria, Flavobacteriales, Flavobacteriaceae, Ulvibacter 3.5 Proteobacteria, Alphaproteobacteria, Rhodobacterales, Rhodobacteraceae 2.7 Bacteroidetes, Flavobacteria, Flavobacteriales, Flavobacteriaceae, Polaribacter 2.0 Bacteroidetes, Flavobacteria, Flavobacteriales, Flavobacteriaceae 1.8 Bacteroidetes, Flavobacteria, Flavobacteriales, Cryomorphaceae 1.7 Unknown Bacteria 1.5 Bacteroidetes, Flavobacteria, Flavobacteriales 1.3 Verrucomicrobia, Opitutae 1.2 Proteobacteria, Alphaproteobacteria, Rhodobacterales, Rhodobacteraceae 1.2 Sum: 20.8 Epipelagic Proteobacteria, Alphaproteobacteria, Rhodobacterales, Rhodobacteraceae 5.6 Unknown Bacteria 3.0 Bacteroidetes, Flavobacteria, Flavobacteriales, Flavobacteriaceae 2.0 Bacteroidetes, Flavobacteria, Flavobacteriales, Flavobacteriaceae 2.0 Bacteroidetes, Flavobacteria, Flavobacteriales, Flavobacteriaceae 1.8 Cyanobacteria, Cyanobacteria, Family II, GpIIa 1.7 Bacteroidetes, Flavobacteria, Flavobacteriales, Flavobacteriaceae 1.6 Proteobacteria, Gammaproteobacteria 1.5 Bacteroidetes, Flavobacteria, Flavobacteriales, Flavobacteriaceae 1.2 Bacteroidetes, Flavobacteria, Flavobacteriales, Flavobacteriaceae 1.2 Sum: 21.7 Mesopelagic Proteobacteria, Alphaproteobacteria, Rickettsiales. SAR11, Pelagibacter 14.0 Proteobacteria 1.5 Proteobacteria, Alphaproteobacteria 1.2 Bacteroidetes, Flavobacteria, Flavobacteriales, Flavobacteriaceae, 1.1 Proteobacteria, Gammaproteobacteria 1.0 Bacteroidetes, Flavobacteria, Flavobacteriales, Flavobacteriaceae, Polaribacter 0.9 Bacteroidetes, Flavobacteria, Flavobacteriales, Flavobacteriaceae 0.9 Unknown Bacteria 0.7 Bacteroidetes, Flavobacteria, Flavobacteriales 0.6 Proteobacteria, Alphaproteobacteria, Rickettsiales. SAR11, Pelagibacter 0.4 Sum: 22.2 Shelf Bottom Proteobacteria, Gammaproteobacteria 16.4 Proteobacteria, Gammaproteobacteria 3.2 Proteobacteria, Gammaproteobacteria 2.2 Proteobacteria, Gammaproteobacteria 2.2 Unknown Bacteria 1.1 Bacteroidetes, Flavobacteria, Flavobacteriales, Flavobacteriaceae 1.0 Proteobacteria, Gammaproteobacteria 1.0 Actinobacteria 0.8 Bacteroidetes 0.7 Bacteroidetes 0.7 Sum: 29.3 Slope Bottom Unknown Bacteria 8.4 Unknown Bacteria 4.2 Bacteroidetes, Flavobacteria, Flavobacteriales, Flavobacteriaceae 2.7 Proteobacteria, Alphaproteobacteria, Rickettsiales. SAR11, Pelagibacter 1.8 Proteobacteria, Gammaproteobacteria 1.8 Proteobacteria, Alphaproteobacteria, Rickettsiales. SAR11, Pelagibacter 1.7 Proteobacteria, Gammaproteobacteria 1.3 Bacteroidetes, Flavobacteria, Flavobacteriales, Flavobacteriaceae 1.2 Proteobacteria, Alphaproteobacteria, Rickettsiales. SAR11, Pelagibacter 1.1 Unknown Bacteria 0.9 Sum: 25.1