FEMS Microbiology Ecology Advance Access published April 1, 2015

Title: Size-fractionated diversity of eukaryotic microbial communities in the

1 2

Eastern Tropical North Pacific oxygen minimum zone

3

Running Title: ETNP protist communities

4

Authors: 1Duret, M.T., 2Pachiadaki, M.G., 3Stewart, F.J., 3Sarode, N., 1Christaki,

5

U., 1Monchy, S., 2*Edgcomb, V.P.

6

1

7

LOG 8187, 32 Ave. FOCH, 62930 Wimereux, France. Current address: Ocean and

8

Earth Science, National Oceanography Centre Southampton, University of

9

Southampton

Laboratoire d’Océanologie et de Géosciences, Université du Littoral, CNRS-UMR

Department of Geology and Geophysics, Woods Hole Oceanographic Institution,

10

2

11

Woods Hole, MA 02543, USA

12

3

13

30332, USA

School of Biology, Georgia Institute of Technology, 310 Ferst Drive, Atlanta, GA

14 15 16

*

17

Geology and Geophysics Department, Woods Hole Oceanographic Institution,

18

Woods Hole, MA, 02543. Tel: 508-289-3734, email: [email protected]

Corresponding author: Dr. Virginia Edgcomb, MS#8, 220 McLean Laboratory,

19

1

19 20

Abstract: Oxygen Minimum Zones (OMZs) caused by water column stratification

21

appear to expand in parts of the world’s ocean, with consequences for marine

22

biogeochemical cycles. OMZ formation is often fueled by high surface primary

23

production, and sinking organic particles can be hotspots of interactions and activity

24

within microbial communities. This study investigated the diversity of OMZ protist

25

communities in two biomass size fractions (>30 and 30-1.6µm filters) from the

26

world's largest permanent OMZ in the Eastern Tropical North Pacific. Diversity was

27

quantified via Illumina MiSeq sequencing of V4 region of 18S SSU rRNA genes in

28

samples spanning oxygen gradients at two stations. Alveolata and Rhizaria

29

dominated the two size fractions at both sites along the oxygen gradient. Community

30

composition at finer taxonomic levels was partially shaped by oxygen concentration,

31

as communities associated with vs. anoxic waters shared only ~32% of OTU (97%

32

sequence identity) composition. Overall, only 9.7% of total OTUs were recovered at

33

both stations and under all oxygen conditions sampled, implying structuring of the

34

eukaryotic community in this area. Size-fractionated communities exhibited different

35

taxonomical features (e.g. Syndiniales Group I in the 1.6-30µm fraction) that could be

36

explained by the microniches created on the surface-originated sinking particles.

37 38

Keywords: Protist diversity, 18S SSU rRNA, Particle-associated, Water-column

2

1 2

Introduction Oxygen Minimum Zones (OMZs) are low oxygen oceanic regions that occur

3

naturally in coastal and open-ocean mesopelagic waters. They are especially

4

prominent in the northern Indian Ocean (Ulloa et al. 2013) and in areas of nutrient

5

upwelling in the Eastern Tropical and Subarctic Pacific (Stramma et al. 2010; Wyrtki

6

1962). Seasonal upwelling delivers cold, oxygenated and nutrient-rich waters to the

7

surface (Stramma et al. 2010), which leads to high primary production in the

8

euphotic layer. A portion of the Particulate Organic Matter (POM) originating from

9

the euphotic zone sinks into the OMZ (i.e., under the surface mixed layer (O2 >

10

180µM; 40 m), Boyd & Trull 2007; Buesseler & Boyd 2009) either directly or indirectly

11

via attachment to decaying or living phytoplankton cells, aggregates, zooplankton

12

feces, etc.; see Alldredge & Silver 1988 and Simon et al. 2002). Sinking POM

13

aggregates fuel aerobic microbial respiration, contributing to a progressive loss of

14

deep water O2 and release of CO2 (reviewed in Simon et al. 2002). OMZs can occur in

15

near-surface waters down to depths greater than 1500 m due to sinking of surface

16

primary production (Kamykowski & Zentara 1990). Depletion of dissolved oxygen

17

can also occur seasonally in coastal and estuarine environments (e.g., Framvaren

18

Fjord, Behnke et al. 2006; Saanich Inlet, Zaikova et al. 2010) or can be a permanent

19

feature of certain anoxic/sulfidic basins (e.g., Cariaco Basin, Taylor et al. 2001). The

20

three major OMZs (Eastern Tropical South Pacific; ETSP, Eastern Tropical North

21

Pacific; ETNP, Arabian Sea) are important sites of denitrification (Codispoti et al.

3

22

2001), OMZs account for about 50% of global oceanic nitrous oxide release to the

23

atmosphere (Keeling et al. 2010). The ETNP (0–25°N; 75–180°W) includes the largest

24

permanent OMZ, representing 41% of the surface area of all global ocean OMZs

25

(~12.106 km², Paulmier & Ruiz-Pino 2009) with an oxygen concentration < 30nM at its

26

core (Tiano et al. 2014). Formation of the ETNP OMZ is mainly driven by a major

27

upwelling system of the North Pacific Ocean that intensifies during the summer

28

months (Wyrtki 1962; Kamykowski & Zentara 1990).

29

Suboxic and anoxic environments (< 20µM and undetectable O2, respectively;

30

Kamykowski & Zentara 1990; Wright et al. 2012) exhibit unique prokaryotic and

31

eukaryotic microbial communities, as shown in studies of OMZs (Stevens & Ulloa

32

2008; Stewart et al. 2012; Orsi et al. 2012) and anoxic basins (Lin et al. 2008; Edgcomb

33

et al. 2011; Orsi et al. 2011). Through microbial denitrification and anaerobic

34

ammonium oxidation processes (anammox, Paulmier & Ruiz-Pino 2009; Canfield et

35

al. 2010; Ulloa et al. 2012), prokaryotic communities in oxygen-depleted waters

36

impact the release of nitrogen gas from ocean waters. While the ecology of

37

prokaryotes is starting to be understood in OMZs, a lot remains to be explored about

38

protist communities in these environments. There is a known shift between

39

eukaryotic communities adapted to aphotic oxic conditions, and those adapted to

40

aphotic anoxic conditions within marine water columns showing steep oxygen

41

gradients (Edgcomb et al. 2011; Orsi et al. 2011, 2012; Stoeck et al. 2003). Symbioses

42

between Bacteria, Archaea and Eukarya in oxygen-depleted water columns appear to

4

43

be common, and may represent a means for eukaryotic protists to exploit otherwise

44

inhospitable habitats. Based on the abundances of putative protist hosts, which

45

frequently harbor hundreds or thousands of prokaryotic partners per host (e.g.,

46

Bernhard et al. 2000; Edgcomb et al. 2011; Orsi et al. 2012), the numerical abundance

47

of prokaryotes in symbiosis with protists may rival that of free-living prokaryotes in

48

some habitats, and the direct impacts of these symbionts on major nutrient

49

biogeochemical cycles may be significant. Sinking POM constitutes a “hotspot” for

50

carbon and nitrogen cycling (Arıstegui et al. 2009; Azam 1998) as it harbors high

51

concentrations of heterotrophic communities (e.g., Cho & Azam 1988; Turley &

52

Mackie 1994). Indeed, heterotrophic protists graze on prokaryotes actively involved

53

in remineralization of POM, and thus modify their quantity, activity, and/or their

54

physiological state (Sherr & Sherr 2002; Frias-Lopez et al. 2009; e.g., regulation of the

55

chemoautotrophic denitrifier Sulfurimonas sp by ciliophorans, Anderson et al. 2013).

56

Protists can also impact nutrient availability by directly modifying and/or

57

remineralizing POM (Taylor 1982; Jumars et al. 1989; Sherr & Sherr 2002). However,

58

little is known about partitioning of eukaryotic communities, between free-living

59

protists and those associated with sinking organic matter, along the oxygen gradient

60

of the OMZ water column.

61

Only one study so far (Parris et al. 2014) has examined OMZ protist

62

communities within multiple biomass size fractions (0.2-1.6µm and > 1.6µm). Taking

63

into account that most of the free-living marine protists have a size of 2.0 to 20µm

5

64

(for heterotrophic nanoflagellates, ciliates and dinoflagellates) we chose to examine

65

two size fractions 1.6-30µm and > 30µm to estimate the community diversity of

66

protists in the free-living and particle-associated fractions, respectively. We focus on

67

samples spanning the vertical oxygen gradient at two locations within the ETNP

68

(continental shelf vs. continental slope). This study contributes to a greater

69

understanding of protist diversity and distribution in the ETNP, along oxygen

70

gradients, and between size fractions.

71

Material and Methods

72

Sample collection. Sample collection took place in June 2013 aboard the R/V New

73

Horizon during the NH1313 cruise. Two sites in the ETNP OMZ off Colima, Mexico

74

were sampled: station 6 (18.55°N, 104.53°W) on the continental shelf, and station 10

75

(18.48°N, 105.12°W) off the continental slope (Fig.1). Samples were collected from the

76

following: the surface mixed layer (O2 concentration > 180µM; 30 m), upper oxycline

77

(~2µM O2; 85 m), and OMZ core (undetectable O2; 100, 125, 300 m) at station 6 (Fig.2-

78

A); the surface mixed layer (O2 > 180µM; 40 m), upper oxycline (~2µM O2; 80 m),

79

OMZ core (undetectable O2; 125, 300, 500, 800 m), and lower oxycline (~5µM O2; 1000

80

m) at station 10 (Fig.2-B). For each depth, microorganisms were collected onto a

81

30µm pore-size nylon mesh filter (47mm) and a 1.6µm nominal pore-size glass fiber

82

filter (GF/A; 47mm) via in-line filtration of roughly 10L of seawater. The larger filter

83

was used to capture free-living microplankton and large POM-attached organisms

84

and the GF/A filter to capture free-living and smaller POM-attached nanoplankton.

6

85

Immediately after filtration, each filter was folded and stored in a cryotube with

86

approximately 2.0ml of extraction buffer (40mM EDTA, 50mM TRIS pH 8.3, 0.73M

87

sucrose) and flash frozen for storage at -80°C until further processing. Two replicates

88

were collected per depth. Oxygen concentration, temperature and salinity were

89

measured with a CTD equipped with a SBE43 dissolved oxygen sensor (Fig.2).

90

Nitrogen oxide concentrations were determined using frozen samples, which were

91

returned to the laboratory and analyzed according to standard protocol by nitrate to

92

nitrite reduction in hot acidic vanadium chloride for colorimetric analysis (Braman &

93

Hendrix 1989; Fig.2).

94

DNA extraction. Samples (n= 48) in extraction buffer were heated to 55°C for 30 min

95

after thawing. Lysates were then purified using a phenol-chloroform-isoamyl alcohol

96

(25:24:1) extraction, precipitated with 100% isopropanol and washed with 70%

97

ethanol. The pelleted DNA was diluted in 50µL of water and kept at -80°C until

98

further processing.

99

DNA amplification and sequencing. The V4 region of 18S SSU rRNA gene was

100

amplified by PCR using the following combination of primers: TAReuk454FWD1 (5′-

101

CCAGCA(G/C)C(C/T)GCGGTAATTCC-3′) and TAReukREV3 (5′-

102

ACTTTCGTTCTTGAT(C/T)(A/G)A-3′) which target most eukaryotes, with some

103

exceptions within the excavates and the microsporidia (Stoeck et al. 2010). Each

104

library was generated using a unique combination per sample of barcoded Illumina

7

105

MiSeq forward and reverse primers designed according to Kozich et al. (2013). Each

106

PCR mix (30µL of MilliQ water, 50µL of buffer [5X], 10µL of MgCl2 [25mM], 5µL of

107

dNTPs [10mM] and 0.5µL of taq polymerase [5u/µL]) contained 1µL of template

108

DNA with 1µL of 10µM forward and reverse primers. The following PCR conditions

109

were used: 5 min at 95°C followed by 40 cycles of 95°C for 30 sec, 55°C for 45 sec,

110

72°C for 60 sec and a final extension of 7 min at 72°C using a Vapo.protect

111

thermocycler (Eppendorf, Germany).

112

Nested PCR was required to amplify the 800 (OMZ core) and 1000 m (lower

113

oxycline) depth samples from station 10. The first PCR conditions were the

114

following: 5 min at 95°C followed by 30 cycles of 95°C for 60 sec, 55°C for 1 min, 72°C

115

for 90 sec followed by a final elongation step of 7 min at 72°C using the forward

116

primer Euk A (5' ATCTGGTTGATYCTGCCAG 3') and Euk B (5’

117

TGATCCTTCTGCAGGTTCACCTAC 3’), amplifying the full rRNA gene (~2kb

118

fragments; Medlin et al. 1988). One microliter of the PCR product was used as a

119

template for a second PCR using the MiSeq primers with the conditions mentioned

120

above (Stoeck et al. 2010).

121

Success of amplification, intensity of the bands and size of the resulting PCR

122

products were checked using gel electrophoresis (1% agarose). Successful amplicons

123

were gel purified using the Zymoclean Gel DNA Recovery Kit (Zymo Research,

124

USA) according to the manufacturer's instructions. The concentration of purified

125

DNA was measured with a Qubit fluorometer (Thermo Fisher Scientific Inc., USA)

8

126

according to the manufacturer’s instructions. Purified amplicons were then pooled

127

in equimolar concentration (4nM) for sequencing on an Illumina MiSeq instrument

128

(School of Biology, Georgia Institute of Technology) running MiSeq control software

129

v2.4.0.4 and using a MiSeq 500 cycle v2 reagent kit. Demultiplexing and preliminary

130

adapter trimming of the samples was carried out using the MiSeq instrument.

131

Bioinformatics analysis. The base quality and basic statistics of demuliplexed read

132

pairs were visualized using FastQC v0.10.1 (Andrews; Babraham Bioinformatics).

133

Reads were screened for Illumina adapters

134

(AATGATACGGCGACCACCGAGATCTACAC and

135

CAAGCAGAAGACGGCATACGAGAT) and quality trimmed using thresholds of a

136

Phred33 score of Q25 and a minimum read length of 100bp with TrimGalore! v0.3.3

137

(Krueger; Babraham Bioinformatics). Pairs containing one mate with an average

138

quality score < 25 or with a length < 100bp after trimming step were discarded. High

139

quality read pairs were merged using FLASH v1.2.9 (Magoč & Salzberg 2011) with

140

the following parameters: average read length (~250bp), fragment length (~400 ±

141

40bp) and a minimum overlap of 20bp between reads. OTU clustering was

142

performed using QIIME v1.8 (Caporaso et al. 2010) with the UCLUST algorithm

143

(Edgar 2010) with a minimum sequence similarity of 97%. The most abundant

144

sequence of each OTU was compared against the Silva V111 Reference Set database

145

(Quast et al. 2013) with the BLAST taxonomic affiliation algorithm (Buhler et al. 2007)

146

on QIIME using the default settings. Non-affiliated OTUs, and those affiliated with

9

147

Metazoa, Archaea, Bacteria and land plants were removed from the data set.

148

Singletons and OTUs whose representatives accounted for less than 0.002% of the

149

overall dataset reads were not used in the subsequent analysis.

150

Statistical analysis. Statistical analyses were computed with R statistics software (R

151

Development Core Team 2008). Reads abundances were normalized using the

152

Hellinger transformation (Legendre & Legendre 2012) in order to minimize the

153

influence of uneven sequencing effort between libraries. Abundances from replicate

154

filters were pooled based on the average of Hellinger transformed values. A

155

Canonical Correspondence Analysis (CCA; Gotelli & Ellison 2004) was used to

156

elucidate relationships between eukaryotic community structure and environmental

157

parameters and an ANOVA was then used to statistically test the influence of these

158

parameters on OTU distributions. Similarities between libraries were visualized with

159

a hierarchical clustering analysis using the Unweighted Pair Group Method with

160

Arithmetic Mean (UPGMA; Legendre & Legendre 2012) based on Bray Curtis

161

dissimilarity. Shared OTU composition between replicates, filter sizes, oxygen

162

conditions and sites was quantified based on the proportion of shared OTUs.

163

Results

164

DNA libraries. A total of 4,227,913 18S rRNA paired-end reads were recovered from

165

the 48 samples analyzed. An uneven number of raw reads was recovered per library

166

(μ= 88082 ± 78745 (se) reads/sample) and this can be explained by possible errors in

167

estimating DNA concentration in some samples prior to pooling for sequencing. This 10

168

would lead to a DNA pool that did not represent equimolar concentrations for all

169

libraries. The effect of this disparity was minimized with the use of Hellinger

170

transformed data. After removal of non-protist OTUs and OTUs representing
30µm fraction at 100 m (OMZ core) of station 6

215

displayed different traits (Supp. Material 2). In the sample from 100 m,

216

Archaeplastida represented 66.1% of total reads, almost all of which were affiliated to

217

Dunaliella sp. We interpret the taxonomic assignment of this result with caution.

218

PCR controls did not suggest any contamination, so the presence of sequences

219

affiliated to Dunaliella sp. in the > 30 µm fraction of this one sample suggests

220

entrainment of intact DNA of this photosynthetic species or an undescribed close

221

relative within a sinking large particle. Stramenopiles were mainly represented in the

222

larger size fraction at both stations (Supp. Material 2 and 3).

223

At finer taxonomic levels, the orders and classes within Alveolata and Rhizaria

224

differed in relative abundances among oxygen conditions and size fractions (Fig.4

225

and 5). Gymnodyniales (38.1 ± 11.8%; Fig.4) were the dominant Alveolates in the

226

larger size fraction at both stations, except at 800 (OMZ core) and 1000 m (lower

227

oxycline) at station 10 which contained a higher proportion of Syndiniales (Fig.4-B).

228

However, some patterns were station-specific, such as the presence of Gonyaulacales

229

in the mixed layer of the column above the continental shelf (Fig.4-A). Syndiniales

13

230

mainly represented Alveolata in the smaller size fraction (44.4 ± 17.9%; Fig.4).

231

Gymnodiniales also represented an important proportion of alveolate richness in all

232

samples (30.0 ± 15.9 %; Fig.4).

233

Orders affiliated with Rhizaria exhibited more complex patterns than those of

234

Alveolates (Fig.5). Both the > 30µm and 1.6-30µm size fractions showed an increased

235

proportion of uncultured Granofilosea with depth (Fig.5). Phaeogromida represented

236

44.6% of Rhizaria sequences in the larger size fractions in the surface mixed layer of

237

both stations (Fig.5-A and B). Collodaria represented 27.5 ± 19.5% of OTUs in the

238

upper oxycline and 100-125 m (OMZ core) at both stations (Fig.5-A and B).

239

Replication. Replicates of the smaller size fraction from station 6 (30, 85, 100, 125 m;

240

surface mixed layer down into the OMZ core) and station 10 (80, 125, 300, 500 m;

241

surface mixed layer down into the OMZ core) clustered in pairs in the UPGMA tree

242

based on a Bray Curtis threshold of 75% similarity. None of the larger size fraction

243

replicates shared 75% or more similarity (Fig.6).

244

Discussion

245

Environmental conditions and community structure. This study investigated the size-

246

fractionated composition of microbial eukaryotic communities in the ETNP OMZ.

247

OMZs harbor a wide diversity of microbial eukaryotic communities showing a non-

248

random partitioning of their members along the oxygen gradient and size fractions

249

(Edgcomb et al. 2011; Orsi et al. 2011, 2012; Ulloa et al. 2012; Parris et al. 2014).

250

Structuring of eukaryotic communities within the ENTP OMZ is evidenced by the 14

251

observation that only 9.7% of the total 2209 OTUs recovered were present in every

252

oxygen condition sampled at both stations. These cosmopolitan OTUs, affiliated

253

predominantly with the taxonomic classes Dinophyceae and Polycistinea, as well as

254

the order Syndiniales, nonetheless showed high variability in abundance (based on

255

rRNA signatures) along the oxygen gradient of the OMZ (Fig.4 and 5). Indeed, the

256

most abundant of these OTUs, a member of Gymnodiniales, was detected at 100-125

257

m (the OMZ core) at both stations, which is consistent with previous reports for its

258

closest relatives in oxygen-depleted habitats (Edgcomb et al. 2011; Orsi et al. 2011,

259

2012). Parasitic group I Syndiniales often accounts for a large majority of sequences

260

in anoxic and suboxic water column samples (Guillou et al. 2008). This was also

261

observed in this study (Fig.4). In contrast, OTUs affiliating with group II Syndiniales

262

were primarily found in the surface mixed layer samples.

263

Radiolarians from Acantharea (Group I, Group VI, Arthracantida; Fig.5) were

264

well represented in this ETNP data set and were detected at every depth regardless

265

of oxygen concentration, as reported previously in the ETSP (Parris et al 2014). This

266

group has been reported to include mixotrophic members, which can produce

267

blooms in surface waters, and are known to host a high number of eukaryotic

268

symbionts (Gilg et al. 2010). These radiolarians are thought to contribute significantly

269

to vertical fluxes of carbon from surface waters due to their large size and their

270

silicate tests that create ballast (Michaels et al. 1995). Nonetheless, radiolarian

271

abundance appeared to vary substantially between depths, as shown for example in

15

272

the smaller size fraction of samples from station 6 (Fig.5-A). OTUs affiliated with

273

Acantharea comprised 96.0% of total OTUs detected in the upper mixed layer and

274

76.7% at 300 m depth (the OMZ core) while representing as few as 12.3% at 100m

275

(top of the OMZ core) at this station. Their uneven abundance along depth is

276

contradictory with the classical vertical flux usually observed for these radiolarians,

277

i.e. an increase in abundance along depth. While estimations of abundances in situ

278

based on relative abundances of rRNA sequences must be interpreted with caution,

279

this may simply reflect periodic pulses of sinking radiolarians coincident with

280

separate bloom events. This hypothesis requires further testing in the future.

281

Oxygen concentration within the ETNP in June 2013 exhibited a typical profile

282

of an upwelling zone (Fernández-Álamo & Färber-Lorda 2006; Stramma et al. 2008),

283

with an OMZ core spanning from 80 to 900 m (Fig.2), while circulation of deep

284

oxygenated waters from the western Pacific along the Equator leads to a deep

285

suboxic to oxic water layer under the Tropical Pacific OMZ (> 1000 m; Fiedler &

286

Talley 2006). Not surprisingly, CCA and ANOVA analyses indicated that O2, along

287

with all the parameters tested, affected eukaryotic community composition (p-value
2µm, Not et al. 2007; 0.2-3µm and 3-10µm, Sauvadet et al. 2010, 0.2-

20

376

1.6 µm, > 1.6 µm, Parris et al. 2014). In these studies, the larger size fractions were

377

dominated by dinoflagellates (Alveolata) and radiolarians (Rhizaria). In the present

378

study, the large free-living predators Gymnodiniales dominated the > 30µm size

379

fraction (Fig.5 and Supp. Material 2 and 3). The 1.6-30µm size fraction was

380

dominated by Synidiales Group I; these groups were also present, at lower levels, in

381

the larger size fraction (Fig.5 and Supp. Material 2 and 3). Described Syndiniales are

382

parasitoids (i.e. the death of their host is necessary for them to complete their life

383

cycle) known to prey on a wide range of species (Guillou et al. 2008), including

384

copepods (Skovgaard et al. 2005). Syndiniales produce small flagellated cells, i.e.,

385

dinospores (from 1 to 12µm) involved in infection transmission to new hosts

386

(Sauvadet et al. 2010). The detection of Syndiniales in the large size fraction most

387

likely corresponds to infected host organisms affiliating with diverse Dinophycea

388

and Rhizaria (Guillou et al. 2008). Likewise, Syndiniales Group I sequences found in

389

the smaller size fraction may originate from infective dinospores dispersed in the

390

water. Alternatively, they may correspond to Syndiniales Group I targeting a smaller

391

host. As parasitoids, Syndiniales could have a large impact on the proportion of

392

POM and dissolved organic matter (DOM) available in the water column through

393

host lysis as well as through modification of host distribution in the water column.

394

Further investigations are needed to understand the role of parasites in marine

395

elemental cycles. The high proportion of Phaeospheria recovered in the large size

396

fraction in the mixed layer of the continental slope station (Fig.5-B) could correspond

397

to a bloom of these polycistineans. The order Collodaria includes radiolarians known 21

398

to play a role in carbon fixation in the oligotrophic ocean, and are the only

399

radiolarians exhibiting a colonial lifestyle (Ishitani et al. 2012). Their sizes and

400

lifestyle explain why these radiolarians were recovered in the larger size fraction of

401

both stations (Fig.5).

402

Ciliates are often prominent members of low oxygen water columns (see review

403

by Edgcomb and Pachiadaki (2014)). While ciliate signatures did not dominate our

404

libraries, examination of taxonomic affiliation within ciliate OTUs shows that OTUs

405

assigned to the oligotrichs Pleuronema and Strombidium were the most common

406

oligotrichs detected in samples from the OMZ cores (the 100 m sample at Station 6

407

and the 125 and 500 m samples at station 10), and were not detected in the oxic

408

mixed layer. Both taxa have been described from suboxic and sulfidic waters of the

409

Black Sea (Wylezich and Jürgens 2011). A similar pattern was observed for other

410

unaffiliated oligotrich OTUs.

411

Phylogenetic differences of the eukaryotic communities existing between the

412

two size fractions could be explained by microniches created by the sinking POM

413

generated in surface waters by primary producers. Mobile protists, such as

414

nanoflagellates, dinoflagellates and ciliates, have the ability to feed on particle-

415

associated bacteria and then detach (Kiørboe et al. 2003). Suspended or sessile protist

416

feeders can also use sinking organic matter aggregates as a ballast to gain access to

417

deeper water layers (ChristensenDalsgaard & Fenchel 2003) rather than feeding

22

418

directly on the aggregates. Sinking particles may therefore carry protists to depths

419

where they may not otherwise inhabit.

420

Replication. Fenchel and Finlay (2004) hypothesized that smaller protists are more

421

likely to have a wider and more homogeneous dispersal. Consistent with this, at both

422

stations, the 1.6-30µm fraction replicates were more similar to one another in terms of

423

composition (49.2 ± 20.9% of shared OTUs; Fig.6-A), compared to replicates in the

424

larger size fraction (25.6 ± 11.4% of shared OTUs; Fig.6-A) at both stations. This may

425

produce a more complete diversity picture for smaller organisms when analyzing the

426

same amount of seawater, as the larger fraction may be less homogeneous.

427

Differences observed between replicates suggest either that ETNP waters are very

428

heterogeneous or that filtration pressures were unequal between replicates.

429

Alternatively, differences in community evenness between the size fractions may

430

affect the potential for replicate sequencing to consistently recover the same OTUs.

431

Conclusion. This study showed that microbial eukaryote communities of the Eastern

432

Tropical North Pacific Oxygen Minimum Zone water column are not only shaped by

433

oxygen concentration, but also by other factors not monitored here. Further

434

descriptive studies have to be carried out in order to better understand factors

435

structuring protist communities of this area (e.g., sulfide or methane concentrations,

436

prokaryotic prey distribution). It would be valuable to collect more information

437

about temporal variations of environmental parameters in this area, in order to

438

understand upwelling currents and their impacts on microbial community dynamics.

23

439

This study also highlights the importance of maintaining the lowest possible

440

pressure drop across filtration membranes in samples in order to separate size

441

fractions more accurately. Despite this, differences in richness and abundances

442

observed between size fractions suggest that, at least partially, taxonomically

443

different protists are associated with POM compared to the free-living communities.

444

Metatranscriptomics studies could provide an understanding of whether particle-

445

attached protist communities are active, as well as identify the nature of their

446

activities.

447

Acknowledgments

448

We thank the captain and crew of the R/V New Horizon. Sample collection was

449

supported by the National Science Foundation (1151698 to FJS). This research was

450

supported by the French Nord Pas de Calais (FRB-DEMO_2013) project and by

451

support to VE from the Woods Hole Oceanographic Ocean Life Institute and Deep

452

Ocean Exploration Institute.

453

References

454 455

Alldredge, A. L. & Silver, M. W. (1988) Characteristics, dynamics and significance of marine snow. Progress in Oceanography 20: 41-82.

456 457 458

Anderson R, Wylezich C, Glaubitz S, Labrenz M & Jürgens K (2013) Impact of protist grazing on a key bacterial group for biogeochemical cycling in Baltic Sea pelagic oxic/anoxic interfaces. Environ Microbiol 15: 1580–1594.

459 460

Arıstegui J, Gasol JM, Duarte CM & Herndl GJ (2009) Microbial oceanography of the dark ocean’s pelagic realm. Limnol Ocean 54: 1501–1529.

461 462

Azam F (1998) Microbial control of oceanic carbon flux: The plot thickens. Science 280: 694–696.

24

463 464 465

Behnke A, Bunge J, Barger K, Breiner H-W, Alla V & Stoeck T (2006) Microeukaryote community patterns along an O2/H2S gradient in a supersulfidic anoxic fjord (Framvaren, Norway). Appl Environ Microbiol 72: 3626–3636.

466 467

Bernhard JM, Buck KR, Farmer MA & Bowser SS (2000) The Santa Barbara Basin is a symbiosis oasis. Nature 403: 77–80.

468 469

Boyd PW & Trull TW (2007) Understanding the export of biogenic particles in oceanic waters: Is there consensus? Prog Oceanogr 72: 276–312.

470 471 472

Braman RS & Hendrix SA (1989) Nanogram nitrite and nitrate determination in environmental and biological materials by vanadium(III) reduction with chemiluminescence detection. Anal Chem 61: 2715–2718.

473 474 475

Buesseler KO & Boyd PW (2009) Shedding light on processes that control particle export and flux attenuation in the twilight zone of the open ocean. Limnol Oceanogr 54: 1210.

476 477 478

Buhler JD, Lancaster JM, Jacob AC & Chamberlain RD (2007) Mercury BLASTN: Faster DNA sequence comparison using a streaming hardware architecture. Proc Reconfigurable Syst Summer Inst.

479 480 481

Canfield DE, Stewart FJ, Thamdrup B, Brabandere LD, Dalsgaard T, Delong EF, Revsbech NP & Ulloa O (2010) A cryptic sulfur cycle in oxygen-minimum–zone waters off the Chilean Coast. Science 330: 1375–1378.

482 483

Caporaso JG, Kuczynski J, Stombaugh J, et al. (2010) QIIME allows analysis of highthroughput community sequencing data. Nat Methods 7: 335–336.

484 485

Cho BC & Azam F (1988) Major role of bacteria in biogeochemical fluxes in the ocean’s interior. Nature 332: 441–443.

486 487 488

Christensen-Dalsgaard, Karen K., and Tom Fenchel (2003) Increased filtration efficiency of attached compared to free-swimming flagellates. Aquatic microbial ecology 33.1: 77-86.

489 490 491

Codispoti LA, Brandes JA, Christensen JP, Devol AH, Naqvi SWA, Paerl HW & Yoshinari T (2001) The oceanic fixed nitrogen and nitrous oxide budgets: Moving targets as we enter the anthropocene? Sci Mar 65: 85–105.

492 493

Edgar RC (2010) Search and clustering orders of magnitude faster than BLAST. Bioinformatics 26: 2460–2461.

494 495 496 497

Edgcomb V, Orsi W, Bunge J, Jeon S, Christen R, Leslin C, Holder M, Taylor GT, Suarez P & Varela R (2011) Protistan microbial observatory in the Cariaco Basin, Caribbean. I. Pyrosequencing vs Sanger insights into species richness. ISME J 5: 1344– 1356.

498 499

Fenchel T (2002) Microbial behavior in a heterogeneous world. Science 296: 1068– 1071.

25

500 501

Fenchel T & Finlay BJ (2004) The ubiquity of small species: Patterns of local and global diversity. BioScience 54: 777–784.

502 503

Fernández-Álamo MA & Färber-Lorda J (2006) Zooplankton and the oceanography of the Eastern Tropical Pacific: a review. Prog Oceanogr 69: 318–359.

504 505

Fiedler PC & Talley LD (2006) Hydrography of the Eastern Tropical Pacific: A review. Prog Oceanogr 69: 143–180.

506 507 508

Frias-Lopez J, Thompson A, Waldbauer J & Chisholm SW (2009) Use of stable isotope-labelled cells to identify active grazers of picocyanobacteria in ocean surface waters. Environ Microbiol 11: 512–525.

509 510

Ganesh S, Parris DJ, DeLong EF & Stewart FJ (2014) Metagenomic analysis of sizefractionated picoplankton in a marine oxygen minimum zone. ISME J 8: 187–211.

511 512 513 514

Gilg IC, Amaral-Zettler LA, Countway PD, Moorthi S, Schnetzer A & Caron DA (2010) Phylogenetic affiliations of mesopelagic Acantharia and Acantharian-like environmental 18S rRNA genes off the Southern California Coast. Protist 161: 197– 211.

515 516

Gotelli NJ & Ellison AM (2004) A Primer Of Ecological Statistics. 1 edition. Sinauer Associates, Sunderland, Mass.

517 518 519

Guillou L, Viprey M, Chambouvet A, Welsh RM, Kirkham AR, Massana R, Scanlan DJ & Worden AZ (2008) Widespread occurrence and genetic diversity of marine parasitoids belonging to Syndiniales (Alveolata). Environ Microbiol 10: 3349–3365.

520 521 522

Ishitani Y, Ujiié Y, de Vargas C, Not F & Takahashi K (2012) Phylogenetic relationships and evolutionary patterns of the order Collodaria (Radiolaria). PLoS ONE 7: e35775.

523 524 525

Jumars PA, Penry DL, Baross JA, Perry MJ & Frost BW (1989) Closing the microbial loop: dissolved carbon pathway to heterotrophic bacteria from incomplete ingestion, digestion and absorption in animals. Deep Sea Res Part I 36: 483–495.

526 527

Kamykowski D & Zentara S-J (1990) Hypoxia in the world ocean as recorded in the historical data set. Deep Sea Res Part I 37: 1861–1874.

528 529 530

Karl, David M., and George A. Knauer (1984) Vertical distribution, transport, and exchange of carbon in the northeast Pacific Ocean: evidence for multiple zones of biological activity. Deep Sea Res Part I 31.3: 221-243.

531 532

Keeling RF, Körtzinger A & Gruber N (2010) Ocean deoxygenation in a warming world. Annu Rev Mar Sci 2: 199–229.

533 534 535

Kiørboe T, Tang K, Grossart H-P & Ploug H (2003) Dynamics of microbial communities on marine snow aggregates: Colonization, growth, detachment, and grazing mortality of attached bacteria. Appl Environ Microbiol 69: 3036–3047.

536 537

Kozich JJ, Westcott SL, Baxter NT, Highlander SK & Schloss PD (2013) Development of a dual-index sequencing strategy and curation pipeline for analyzing amplicon 26

538 539

sequence data on the MiSeq Illumina sequencing platform. Appl Environ Microbiol 79: 5112–5120.

540

Legendre P & Legendre LFJ (2012) Numerical Ecology. Elsevier.

541 542

Lin X, Scranton MI, Chistoserdov AY, Varela R & Taylor GT (2008) Spatiotemporal dynamics of bacterial populations in the anoxic Cariaco Basin. Limnol Oceanogr 53: 37.

543 544

Magoč T & Salzberg SL (2011) FLASH: fast length adjustment of short reads to improve genome assemblies. Bioinformatics 27: 2957–2963.

545 546

Medlin L, Elwood HJ, Stickel S & Sogin ML (1988) The characterization of enzymatically amplified eukaryotic 16S-like rRNA-coding regions. Gene 71: 491–499.

547 548 549

Michaels AF, Caron DA, Swanberg NR, Howse FA & Michaels CM (1995) Planktonic sarcodines (Acantharia, Radiolaria, Foraminifera) in surface waters near Bermuda: Abundance, biomass and vertical flux. J Plankton Res 17: 131–163.

550 551

Not F, Gausling R, Azam F, Heidelberg JF & Worden AZ (2007) Vertical distribution of picoeukaryotic diversity in the Sargasso Sea. Environ Microbiol 9: 1233–1252.

552 553

Not F, del Campo J, Balagué V, de Vargas C & Massana R (2009) New insights into the diversity of marine picoeukaryotes. PLoS ONE 4: e7143.

554 555 556

Orsi W, Edgcomb V, Jeon S, Leslin C, Bunge J, Taylor GT, Varela R & Epstein S (2011) Protistan microbial observatory in the Cariaco Basin, Caribbean. II. Habitat specialization. ISME J 5: 1357–1373.

557 558

Orsi W, Song YC, Hallam S & Edgcomb V (2012) Effect of oxygen minimum zone formation on communities of marine protists. ISME J 6: 1586–1601.

559 560 561

Parris DJ, Ganesh S, Edgcomb VP, DeLong EF & Stewart F (2014) Microbial eukaryote diversity in the marine oxygen minimum zone off northern Chile. Aquat Microbiol 5: 543.

562 563

Paulmier A & Ruiz-Pino D (2009) Oxygen minimum zones (OMZs) in the modern ocean. Prog Oceanogr 80: 113–128.

564 565 566

Quast C, Pruesse E, Yilmaz P, Gerken J, Schweer T, Yarza P, Peplies J & Glockner FO (2013) The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic Acids Res 41: D590–D596.

567 568 569

R Development Core Team (2008) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria http://www.Rproject.org.

570 571 572 573

Saito, K., Drgon, T., Robledo, J. A., Krupatkina, D. N., & Vasta, G. R. (2002). Characterization of the rRNA locus of Pfiesteria piscicida and development of standard and quantitative PCR-based detection assays targeted to the nontranscribed spacer. Appl Environ Microbiol 11: 5394-5407.

27

574 575 576

Sauvadet A-L, Gobet A & Guillou L (2010) Comparative analysis between protist communities from the deep-sea pelagic ecosystem and specific deep hydrothermal habitats. Environ Microbiol 12: 2946–2964.

577 578

Sherr EB & Sherr BF (2002) Significance of predation by protists in aquatic microbial food webs. Antonie Van Leeuwenhoek 81: 293–308.

579 580

Simon M, Grossart H-P, Schweitzer B & Ploug H (2002) Microbial ecology of organic aggregates in aquatic ecosystems. Aquat Microb Ecol 28: 175–211.

581 582 583

Skovgaard A, Massana R, Balagué V & Saiz E (2005) Phylogenetic position of the copepod-infesting parasite Syndinium turbo (Dinoflagellata, Syndinea). Protist 156: 413–423.

584 585

Stevens H & Ulloa O (2008) Bacterial diversity in the oxygen minimum zone of the Eastern Tropical South Pacific. Environ Microbiol 10: 1244–1259.

586 587

Stewart FJ, Ulloa O & DeLong EF (2012) Microbial metatranscriptomics in a permanent marine oxygen minimum zone. Environ Microbiol 14: 23–40.

588 589

Stoeck T, Taylor GT & Epstein SS (2003) Novel eukaryotes from the permanently anoxic Cariaco Basin (Caribbean Sea). Appl Environ Microbiol 69: 5656–5663.

590 591 592

Stoeck T, Bass D, Nebel M, Christen R, Jones MDM, Breiner H-W & Richards TA (2010) Multiple marker parallel tag environmental DNA sequencing reveals a highly complex eukaryotic community in marine anoxic water. Mol Ecol 19: 21–31.

593 594

Stramma L, Johnson GC, Sprintall J & Mohrholz V (2008) Expanding oxygenminimum zones in the Tropical Oceans. Science 320: 655–658.

595 596

Stramma L, Schmidtko S, Levin LA & Johnson GC (2010) Ocean oxygen minima expansions and their biological impacts. Deep Sea Res Part I 57: 587–595.

597 598

Taylor GT (1982) The role of pelagic heterotrophic protozoa in nutrient cycling: a review. Ann Inst Ocean Paris 58: 227–241.

599 600 601 602

Taylor TG, Iabichella M, Ho T-Y, Scranton IM, Thunell CR, Muller-Karger F & Varela R (2001) Chemoautotrophy in the redox transition zone of the Cariaco Basin: A significant midwater source of organic carbon production. Limnol Oceanogr 46: 148– 163.

603 604 605

Tiano L, Garcia Robledo E, Dalsgaard T, Devol A, Ward BB, Ulloa O, Canfield DE & Peter Revsbech N (2014) Oxygen distribution and aerobic respiration in the North and South Eastern Tropical Pacific oxygen minimum zones. Deep Sea Res Part I.

606 607 608

Turley CM & Mackie PJ (1994) Biogeochemical significance of attached and freeliving bacteria and the flux of particles in the NE Atlantic Ocean. Mar Ecol-Prog Ser 115: 191–191.

609 610

Ulloa O, Canfield DE, DeLong EF, Letelier RM & Stewart FJ (2012) Microbial oceanography of anoxic oxygen minimum zones. Proc Natl Acad Sci 109: 15996–16003. 28

611 612 613

Ulloa PO, Wright JJ, Belmar L & Hallam SJ (2013) Pelagic Oxygen Minimum Zone Microbial Communities. The Prokaryotes, (Rosenberg E, DeLong EF, Lory S, Stackebrandt E & Thompson F, eds), pp. 113–122. Springer Berlin Heidelberg

614 615

Wright JJ, Konwar KM & Hallam SJ (2012) Microbial ecology of expanding oxygen minimum zones. Nat Rev Microbiol 10: 381–394.

616 617

Wyrtki K (1962) The oxygen minima in relation to ocean circulation. Deep Sea Res Oceanogr Abstr 9: 11–23.

618 619 620 621

Zaikova E, Walsh DA, Stilwell CP, Mohn WW, Tortell PD & Hallam SJ (2010) Microbial community dynamics in a seasonally anoxic fjord: Saanich Inlet, British Columbia. Environ Microbiol 12: 172–191.

29

621

622 623

Figure 1 - Sampling map. Two sites were sampled for genomics in the ETNP, off the

624

coast of Colima (Mexico) along the oxygen gradient within depths: five depths in

625

station 6 (maximal depth: 1500m; 18.55°N, 104.53°W), and seven depths in station 10

626

(maximal depth: 2600 m; 18.48°N, 105.12°W).

627 628

30

A

0

50 50

Oxygen (µM) 100 100 150

150200

200 250

0

Depth (m) Depth (m)

200 0

400 200 600 400 800 600

1000 800 1200 1000 0

5 10

10

1200 0

10

628

2015 2030 25 Temperature Nitrite + Nitrate(°C) (µM)

40 30

35 50

20 30 Nitrite + Nitrate (µM)

40

50

Fig. 2-A

629

B

0

50 50

Oxygen (µM) 100 100 150

150200

250 200

Depth (m)Depth (m)

0 200 0 400 200 600 400 800 600 1000 800 1200 1000 0

5 10

10

2015 2030 25 Nitrite Temperature + Nitrate(°C) (µM)

40 30

50 35

20 30 Nitrite + Nitrate (µM)

40

50

1200 0 630 631

10

Fig. 2-B

31

X

Oxygen (µM) Temperature (°C) Nitrate + Nitrite (µM) Sampled depth

632 633

Figure 2 - Vertical profiles of environmental parameters. CTD measurements of

634

oxygen concentration (µM; split line), temperature (°C; dashed line) and nitrite +

635

nitrate concentration (µM; dots) along depth in station 6 (A) and station 10 (B) of the

636

ETNP during June 2013. Crosses indicate depths sampled for DNA. Two filter sizes

637

(30 and 1.6µm) in replicates (n= 2) were collected per depth (4 samples/depth).

638

32

638

Mixed Layer 1.0%

A

23.8%

29.9% 18.7%

Upper Oxycline 8.3%

63.3%

OMZ Core 18.7%

639 640

Fig. 3A

641

33

641

B

Upper Oxycline 3.7%

OMZ Core 4.1% 52.3%

79.1% Lower Oxycline 0.3%

41.5%

35.2% 22.5%

18.7% 33.9%

46.1% 20.0%

Surface Oxic 2.7%

26.5%

31.0% 642 643

Fig. 3B

644

Figure 3 - Protist community structure within oxygen gradient. Venn diagrams of

645

shared diversity between oxygen conditions in station 6 (A) and station 10 (B) of the

646

ETNP during June 2013. Percentages of unique OTUs per oxygen condition are

647

relative to the number of total OTUs within each station (n= 1847 and n= 2196 OTUs

648

for station 6 and station 10, respectively). Percentages of shared OTUs are relative to

649

the total number of OTUs present in the compared oxygen conditions. The color

650

indicates the oxygen condition (blue corresponds to the mixed layer (> 180μM O2),

651

orange corresponds to the upper oxycline (~2μM), red corresponds to the OMZ core

652

(~0μM) and purple corresponds to the lower oxycline (~5μM)).

653

34

653

654 655

Fig. 4-A

656

35

656

657 658

Fig. 4-B 36

659 660

Figure 4 - Community alveolates diversity within depth. Diversity within Alveolata

661

from station 6 (A) and station 10 (B) of the ETNP within depth for both size fractions

662

(> 30µm and 1.6-30µm). Replicates were pooled and the average of their Hellinger

663

transformed abundances was used. Oxygen conditions are indicated next to depth

664

(ML: Mixed Layer, UO: Upper Oxycline, OMZ: OMZ core, LO: Lower Oxycline).

665

Taxa represented by less than 0.5% on average per sample were pooled and

666

represented as “others”. Abundance of these specific taxa can be found in Supp.

667

Material 3.

668

37

668

669 670

Fig. 5-A

671

38

671

672 673

Fig. 5-B

39

674 675

Figure 5 - Community rhizarian diversity within depth. Diversity within Rhizaria

676

from station 6 (A) and station 10 (B) of the ETNP within depth for both size fractions

677

(> 30µm and 1.6-30µm). Replicates were pooled and the average of their Hellinger

678

transformed abundances was used. Oxygen conditions are indicated next to depth

679

(ML: Mixed Layer, UO: Upper Oxycline, OMZ: OMZ core, LO: Lower Oxycline).

680

Taxa represented by less than 0.5% on average per sample were pooled and

681

represented as “others”. Abundance of these specific taxa can be found in Supp.

682

Material 4.

683

40

683

A 0%#

Shared#diversity#(%#of#OTUs)# 20%# 40%# 60%# 80%#

Depth#(m)#(Oxygen#condiGon)#

40#(ML)#

100%#

*

80#(UO)#

* * *

#100#(OMZ)# 125#(OMZ)#

>#30µm#J#StaGon#6# >#30µm#J#StaGon#10#

**

300#(OMZ)#

1.6J30µm#J#StaGon#6# 1.6J30µM#J#StaGon#10#

*

500#(OMZ)#

*

800#(OMZ)# 1000#(LO)#

684 685

Fig. 6-A Bray-Curtis dissimilarity

B 0.0

0.2

0.4

0.6

0.8

1.0

x1 10 A 125 x2 10 B 40 x1 10 B 40 x1 6 B 30 x2 6 B 30 x2 6 B 300 x2 10 A 40 x2 6 A 30 x1 6 A 30 x1 10 A 40 x2 6 A 300 x1 6 A 300 x1 10 A 80 x2 10 A 80 x2 6 A 85 x1 6 A 85 x2 10 B 80 x1 10 B 80 x2 6 B 85 x1 6 B 85 x2 6 A 125 x2 6 A 100 x2 10 A 500 x1 6 A 125 x1 10 A 300 x2 10 A 125 x1 10 A 500 x2 10 A 300 x1 6 B 300 x2 10 B 300 x1 10 B 300 x2 10 B 500 x1 10 B 500 x2 6 B 100 x1 6 B 100 x2 10 B 125 x1 10 B 125 x1 6 B 125 x2 6 B 125 x2 10 A 800 x2 10 B 800 x2 10 A 1000 x2 10 B 1000 x1 10 B 800 x1 10 B 1000 x1 10 A 800 x1 10 A 1000 x1 6 A 100

!"

686 687

Fig. 6-B 41

688 689 690 691 692 693 694 695 696 697 698 699 700

Figure 6 - Shared diversity between replicates and similarity clusters. (A) Proportion of shared OTUs between replicates of the > 30µm fraction (blue) the 1.6-30µm fraction (red) in station 6 and station 10 of the ETNP within depth. Red asterisks show replicates clustered in pairs at a 75% threshold similarity level. Oxygen conditions are indicated next to depth (ML: Mixed Layer, UO: Upper Oxycline, OMZ: OMZ core, LO: Lower Oxycline). (B) UPGMA hierarchical tree based Bray Curtis dissimilarities between libraries from station 6 and station 10. Red frame show libraries clustered together at a 75% threshold similarity level. The color indicates the oxygen condition (blue corresponds to the mixed layer (> 180μM O2), orange corresponds to the upper oxycline (~2μM), red corresponds to the OMZ core (~0μM) and purple corresponds to the lower oxycline (~5μM)). The sample names follow this logic: x1/2 (corresponds to the replicate number), A/B (corresponds to the filter size: A - 30µm and B – 1.6µm), depth in m.

701

42