Environment-Dependent Distribution of the Sediment nifH-HarboringMicrobiota in the Northern South China Sea

Hongyue Dang,a,b Jinying Yang,a Jing Li,a Xiwu Luan,c,d Yunbo Zhang,e Guizhou Gu,e Rongrong Xue,e Mingyue Zong,e

Martin G. Klotza,f

State Key Laboratory of Heavy Oil Processing, Key Laboratory of Bioengineering and Biotechnology in Universities of Shandong, Centre for Bioengineering andBiotechnology, China University of Petroleum (East China), Qingdao, Chinaa; State Key Laboratory of Marine Environmental Science, Xiamen University, Xiamen, Chinab;Key Laboratory of Marine Hydrocarbon Resources and Environmental Geology, Ministry of Land and Resources of China, Qingdao, Chinac; Qingdao Institute of MarineGeology, Qingdao, Chinad; College of Chemical Engineering, China University of Petroleum (East China), Qingdao, Chinae; Department of Biology, University of NorthCarolina, Charlotte, North Carolina, USAf

The South China Sea (SCS), the largest marginal sea in the Western Pacific Ocean, is a huge oligotrophic water body with verylimited influx of nitrogenous nutrients. This suggests that sediment microbial N2 fixation plays an important role in the produc-tion of bioavailable nitrogen. To test the molecular underpinning of this hypothesis, the diversity, abundance, biogeographicaldistribution, and community structure of the sediment diazotrophic microbiota were investigated at 12 sampling sites, includ-ing estuarine, coastal, offshore, deep-sea, and methane hydrate reservoirs or their prospective areas by targeting nifH and someother functional biomarker genes. Diverse and novel nifH sequences were obtained, significantly extending the evolutionarycomplexity of extant nifH genes. Statistical analyses indicate that sediment in situ temperature is the most significant environ-mental factor influencing the abundance, community structure, and spatial distribution of the sediment nifH-harboring micro-bial assemblages in the northern SCS (nSCS). The significantly positive correlation of the sediment pore water NH4

� concentra-tion with the nifH gene abundance suggests that the nSCS sediment nifH-harboring microbiota is active in N2 fixation and NH4

�

production. Several other environmental factors, including sediment pore water PO43� concentration, sediment organic carbon,

nitrogen and phosphorus levels, etc., are also important in influencing the community structure, spatial distribution, or abun-dance of the nifH-harboring microbial assemblages. We also confirmed that the nifH genes encoded by archaeal diazotrophs inthe ANME-2c subgroup occur exclusively in the deep-sea methane seep areas, providing for the possibility to develop ANME-2cnifH genes as a diagnostic tool for deep-sea methane hydrate reservoir discovery.

Microbial fixation of nitrogen (N) is an important process inthe global N biogeochemical cycle (1, 2), providing bioavail-

able ammonia as the bioenergetically least costly source of nitro-gen, which is especially important in oligotrophic ecosystems. Thecurrent N budget of the marine N cycle is considered unbalancedcaused by N-loss exceeding N-gain. Such an imbalance, if real,would eventually lead to an end of primary production in theoceans, and it has thus been hypothesized that the N-gain due toN2 fixation in the ocean is underestimated (3, 4). Although theprevailing theory assumes that N2 fixation ceases in such environ-ments, recent findings suggest that microbial N2 fixation operatesalso in marine environments served with bioavailable N fromother sources (5–8). Thus, diazotrophic microorganisms may bemore widespread and play more important roles in marine envi-ronments than previously suggested. Although N2 fixation (nitro-genase activity) is performed by numerous groups of microorgan-isms (9) and the nitrogenase reductase-encoding gene, nifH, hasbeen frequently targeted as a functional biomarker for investigat-ing the diazotrophic community in natural environments (10,11), the diversity and community structure of the nifH-harboringmicroorganisms are still poorly understood globally, and manyimportant diazotrophs remain to be discovered (12–18).

Deep-sea sediments constitute the largest fraction of theEarth’s surface and a vast portion of these sediments are anoxicand thus suitable habitat for microbial N2 fixation (19); however,marine diazotrophic microbes have been studied mostly in oligo-trophic surface waters (see reference 20 and references therein)and certain shallow-water benthic systems, such as salt marshes,

mangroves, microbial mats and other coastal environments (21–26). Rarely conducted investigations of the N2 fixation microbialcommunities of deep-sea waters and hydrothermal vent fluidshave discovered diverse and novel nifH gene sequences (12, 13).Although this correlates with the finding that dissolved N2 isabundant (0.59 mM) in the deep ocean (12), these environmentshave been sparsely investigated regarding the diversity and com-munity structure of the diazotrophic microbiota.

Marginal seas are major areas of active biogeochemical cyclingprocesses (27). These areas may absorb a significant fraction ofatmospheric CO2 by high-level primary production and organicmatter export to deep oceans and sediments (28, 29), playing animportant role in alleviating the ongoing global warming and cli-mate change. However, this capability usually does not operate atfull capacity due to the lack of adequate nutrient input for aquaticbioproductivity. Marine N2 fixation plays an important role ingeneral in marine ecosystem and an even more important role in

Received 15 June 2012 Accepted 1 October 2012

Published ahead of print 12 October 2012

Address correspondence to Hongyue Dang, DangHY20042000@yahoo.com.cn.

J.Y. and J.L. contributed equally to this article.

Supplemental material for this article may be found at http://dx.doi.org/10.1128/AEM.01889-12.

Copyright © 2013, American Society for Microbiology. All Rights Reserved.

doi:10.1128/AEM.01889-12

January 2013 Volume 79 Number 1 Applied and Environmental Microbiology p. 121–132 aem.asm.org 121

on March 18, 2021 by guest

http://aem.asm

.org/D

ownloaded from

areas of the tropical and subtropical ocean (30). N2 fixation, alongwith atmospheric N deposition, is the only source of new N thatcan lead to a net sequestration of atmospheric CO2 in the deepocean (31). In the marine benthic ecosystem, recent studies indi-cate that archaea, especially the anaerobic methanotrophic(ANME) archaeal functional guild (32), constitute a great portionof the in situ diazotrophic community and play an important rolein supplying newly fixed nitrogen in deep-sea sediments rich inmethane hydrates (14–17). Due to the global distribution of meth-ane hydrates in marginal sea sediments (4, 33), the ANMEdiazotrophs’ contribution to the marine and global N2 fixationcannot be neglected. However, it is currently still unknownwhether archaeal diazotrophs are specific to methane-charged en-vironments, such as deep-sea sediments rich in methane hydrates,or widely distributed in sediments even without methane hy-drates. The distribution and contribution of archaeal diazotrophsin marine sediments need to be investigated.

The South China Sea (SCS), the largest marginal sea (�3.5 �106 km2) in the western Pacific Ocean, is a huge subtropical andtropical oligotrophic water body with usually undetectable nitrateand phosphate in the euphotic layer (34). Since the seawater av-erage N/P is below the Redfield ratio, productivity in the SCS islikely limited by the supply of nitrogenous nutrients (35). Severallarge rivers, such as the Pearl River and the Mekong River, dis-charge into the SCS. Riverine input may provide nutrients forphytoplankton growth and bioproduction. However, this input islocalized to the estuarine and coastal areas and does not providethe nutrients for the whole oligotrophic SCS (36). Monsoonalwind driving eddies and other upwelling and vertical mixing anddiffusing processes may also provide nutrient supply from deepwater to surface water in the SCS (37, 38). Thus, sediment micro-bial N2 fixation may play an important role in maintaining deep-water nutrient availability and surface water nutrient supply.

Seawater diazotrophic bacteria and N2 fixation potential have

been previously investigated in several areas of SCS (6, 34, 38–42).Sediment diazotrophic microbiota have not yet been investigatedeven though they may play an important role in providing newlyfixed nitrogen to the oligotrophic SCS ecosystem. In addition,evidenced and prospective methane hydrate reservoirs have beenidentified in the northern South China Sea (nSCS) (43–46), andthese areas may sustain productive methane-seep-specific ecosys-tems, requiring a high supply of nitrogenous nutrients and thusfavoring the development of an active microbial diazotroph com-munity. The nSCS provides an ideal ecosystem to investigate thedistribution and contribution of sediment microbial diazotrophsin general, and the archaeal diazotrophs in particular, to the ma-rine N cycling. In the present study, numerous sampling stationswere selected, including estuarine, coastal, offshore, deep-sea, andmethane hydrate reservoirs or their prospective areas to cover themajor environments characteristic to the nSCS to investigate thediversity, abundance, distribution, community composition, andstructure of the sediment diazotrophic microbiota and major en-vironmental factors that influence their ecological features.

MATERIALS AND METHODSSample collection and environmental factor measurements. Sedimentsamples were collected from 12 stations of the nSCS using a 0.1-m2 stain-less steel Gray O’Hara box corer or a deep-sea sediment grab samplerduring an Open Cruise of R/V Shiyan 3 in August of 2007 (Fig. 1). Onlyundisturbed samples with clear overlying seawater were collected to en-sure the integrity of the surface sediment structures. Replicate surfacesediment subcore samples down to a 5-cm depth for microbiological andenvironmental analyses were taken aseptically with sterile 60-ml syringes(luer end removed), homogenized, and stored in airtight sterile plasticbags at �20°C during the cruise and at �80°C after delivery to the labo-ratory.

Surface sediment temperature was measured on deck inside the sam-pler once the sediments were collected. Other environmental factors weremeasured in the laboratory. Sediment organic carbon (OrgC) and organic

FIG 1 Map showing the 12 sediment sampling sites in the nSCS. (Modified from Zhou et al. [47] with permission from Springer Science�Business Media.)

Dang et al.

122 aem.asm.org Applied and Environmental Microbiology

on March 18, 2021 by guest

http://aem.asm

.org/D

ownloaded from

nitrogen (OrgN) contents were measured with a PE 2400 Series IICHNS/O elemental analyzer (Perkin-Elmer, Norwalk, CT). The sedimentcontents of water (WC), total organic matter (OM), total phosphorus(TP), inorganic phosphorus (IP), and organic phosphorus (OrgP) weremeasured according to the method of Danovaro (48). Sediment porewater dissolved nutrient concentrations, such as nitrate (NO3

�), nitrite(NO2

�), ammonium (NH4�), and phosphate (PO4

3�), were measuredwith a nutrient QUAATRO AutoAnalyzer (Bran�Luebbe, Germany).Sediment pore water dissolved urea concentration was measured accord-ing to the method of Grasshoff et al. (49). A Cilas 940L laser granulometer(Company Industrielle des Lasers, France) was used for sediment granu-larity analysis (Table 1).

DNA extraction and nifH gene clone library analyses. Sediment mi-crobial community DNA was extracted and pooled from replicate subcoresamples of each station using a FastPrep DNA extraction kit for soil and aFastPrep-24 cell disrupter (MP Biomedicals, Solon, OH) as described pre-viously (15, 50). DNA concentrations were measured using PicoGreen(Molecular Probes, Eugene, OR) and a Modulus single-tube multimodereader fluorometer (Turner Biosystems, Sunnyvale, CA). Bacterial andarchaeal nitrogenase reductase genes (nifH, including anfH and vnfH)were amplified with the primers nifHfw and nifHrv (12, 15, 17). To testthe reproducibility of our experimental procedure and to identify any

potential small-scale (�20 cm) spatial variability of the sediment di-azotrophic community, two separate nifH gene clone libraries (E407-Iand E407-II) were constructed for sampling station E407, each from adistinct subcore DNA sample. PCR product cloning followed previousprocedures (15, 50). Cloned gene fragments were reamplified to check thecorrect size of the DNA inserts using vector primers M13-D and RV-M(51), which were also used for sequencing with an ABI 3770 sequencer(Applied Biosystems, Foster City, CA). The resultant DNA sequenceswere translated into conceptual NifH protein sequences, and the BLASTpprogram was used for retrieval of the top-hit sequences from GenBank(52). NifH sequences were grouped into operational taxonomic units(OTU) within 0.05 sequence distance calculated by using the DOTURprogram (53). Alignments of the NifH protein sequences, obtained withthe program CLUSTAL X version 2.0 (54), were used for inference ofphylogeny with PHYLIP version 3.69 (55), as reported previously (15).

Quantification of major microbial groups. The abundances of sedi-ment bacteria, archaea, nifH-harboring microorganisms, ANME di-azotrophs, and methanogenic and methanotrophic archaea were quanti-fied by using real-time fluorescent quantitative PCR (qPCR) methodswith group-specific primers targeting, respectively, the 16S rRNA, nifH,nifD (encoding the �-subunit of nitrogenase molybdenum-iron protein),and mcrA (encoding the �-subunit of methyl-coenzyme M reductase)

TABLE 1 Measurements of in situ environmental parameters of the 12 sampling stations in the northern South China Sea

Environmentalfactora

Station

A3 CF6 CF8 CF11 CF14 CF15 E407 E422 E501 E504 E505 E801

Longitude (°E) 114°23.163= 119°30.060= 111°3.710= 114°34.477= 115°12.971= 115°29.373= 120°0.017= 112°0.793= 110°41.835= 111°18.096= 111°29.029= 110°20.410=Latitude (°N) 21°50.525= 22°0.316= 18°1.908= 19°43.341= 19°54.256= 19°59.581= 18°29.810= 18°0.341= 18°59.995= 19°0.102= 18°59.897= 17°11.450=Water depth (m) 52 2,441 1,548 1,050 1,220 1,300 1,800 2,456 65 131 154 1,400

SedimentTemp (°C) 21.9 5.5 5.0 5.2 3.6 3.4 4.7 3.8 20.5 17.7 17.4 3.8WC (%) 36.92 29.66 27.92 32.25 33.19 36.40 32.51 32.37 18.23 35.22 15.22 27.94OM (%) 1.89 3.06 2.26 2.10 1.83 2.68 2.33 2.09 1.08 1.79 1.58 3.22OrgC (%) 1.00 0.63 0.96 1.32 1.32 1.27 0.67 0.94 0.28 0.76 0.47 1.09OrgN (%) 0.10 0.08 0.12 0.16 0.15 0.17 0.06 0.11 0.03 0.08 0.06 0.14OrgP (�mol/g) 5.92 3.52 6.17 8.69 7.62 12.62 8.36 4.28 3.11 6.31 2.55 8.55IP (�mol/g) 19.89 23.52 16.56 17.53 16.56 13.99 16.31 18.64 7.81 17.61 15.33 15.23TP (�mol/g) 25.81 27.04 22.73 26.22 24.18 26.61 24.67 22.92 10.92 23.92 17.88 23.78OrgC/OrgN 10.00 7.88 8.00 8.25 8.80 7.47 11.17 8.55 9.33 9.50 7.83 7.79Sand content (%) 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 38.18 0.14 0.62 0.00Silt content (%) 64.91 63.00 62.93 62.41 62.74 60.75 53.61 59.29 44.46 69.72 75.7 57.94Clay content (%) 35.09 36.97 37.07 37.59 37.26 39.25 46.39 40.71 17.37 30.14 23.68 42.06Median grain size

(ø)7.00 7.27 7.12 7.14 7.15 7.28 7.78 7.34 4.75 6.68 6.12 7.58

Mean grain size(ø)

7.27 7.49 7.42 7.42 7.37 7.48 7.74 7.61 5.43 7.01 6.52 7.71

Sorting coefficient 1.65 1.50 1.59 1.60 1.63 1.60 1.56 1.54 2.25 1.69 1.79 1.40Kurtosis 1.95 1.80 1.86 1.88 1.92 1.88 1.85 1.78 2.74 2.02 2.18 1.65Skewness 0.95 0.84 0.97 0.93 0.87 0.78 –0.76 0.88 1.94 1.17 1.49 0.78

Sediment pore waterSalinity (‰) 31.5 31.2 31.0 30.0 30.5 31.5 31.5 30.0 30.1 31.0 32.5 31.0pH 7.17 7.11 7.40 7.16 7.15 7.15 7.06 7.17 7.28 7.13 7.35 7.06Eh (mv) –15.0 –11.0 –28.0 –14.0 –13.0 –13.0 –8.0 –14.0 –21.0 –12.0 –27.0 –8.0DO (�M) 118.75 203.13 128.13 178.13 237.50 140.63 237.50 234.38 115.63 96.88 165.63 215.63NO3

� (�M) 14.31 3.61 1.46 13.58 14.92 6.43 22.67 17.42 3.95 2.85 6.30 9.05NO2

� (�M) 0.42 0.83 0.60 1.09 1.19 1.00 1.86 1.53 0.63 2.80 0.56 3.99NOx

� (�M)b 14.73 4.44 2.06 14.67 16.11 7.43 24.53 18.95 4.58 5.65 6.86 13.04NH4

� (�M) 47.11 5.97 54.25 8.22 105.94 13.76 126.68 8.72 525.74 59.11 32.70 475.56DIN (�M)c 61.84 10.41 56.31 22.89 122.05 21.19 151.21 27.67 530.32 64.76 39.56 488.60PO4

3� (�M) 5.49 6.26 6.42 7.54 6.19 6.29 8.46 7.62 5.07 9.08 4.42 14.42N/P (DIN/PO4

3�) 11.26 1.66 8.77 3.04 19.72 3.37 17.87 3.63 104.60 7.13 8.95 33.88Urea (�M) 1.91 1.00 1.24 1.16 2.07 2.82 0.58 1.58 2.41 1.33 2.57 5.06

a WC, water content; OM, organic matter; OrgC, organic C; OrgN, organic N; OrgP, organic P; IP, inorganic P; TP, total P; ø, the Krumbein phi scale of grain size; Eh, redoxpotential.b NOx

� was calculated as the sum of NO2� and NO3

�.c The total inorganic N concentration (DIN) was calculated as the sum of NH4

�, NO2�, and NO3

�.

South China Sea Deep-Sea Sediment nifH Diversity

January 2013 Volume 79 Number 1 aem.asm.org 123

on March 18, 2021 by guest

http://aem.asm

.org/D

ownloaded from

genes (see Table S1 in the supplemental material). All qPCR assays werecarried out in triplicate with an ABI Prism 7500 sequence detection system(Applied Biosystems) using the SYBR green qPCR protocols (17, 56).Reaction conditions of qPCR were optimized with reference plasmidscarrying the respective target genes constructed in this and previous stud-ies (15, 51, 56). In all qPCR experiments, fluorescence reads were carriedout at 72 or 80°C, and negative controls lacking template DNA were sub-jected to the same qPCR procedures to detect any possible contaminationor carryover. Agarose gel electrophoresis and melting-curve analysis wereroutinely used to confirm the specificity of the qPCRs. Melting curveswere obtained at 60 to 95°C, with a read every 1°C and holding for 1 sbetween reads. The resultant qPCR data were analyzed with the second-derivative maximum method using the ABI Prism 7500 SDS software(version 1.4; Applied Biosystems) (56, 57).

For the qPCR assays, standard curves were generated by serial dilutionof reference plasmids containing target 16S rRNA, nifH, nifD, or mcrAgene fragments as the insert, with plasmids extracted from E. coli hostsusing a Mini plasmid kit (Qiagen, Valencia, CA) and linearized with anendonuclease specific for the vector region. Concentrations of linearizedplasmid DNAs were measured using PicoGreen and a Modulus single-tube multimode reader fluorometer. The ranges of the reference plasmidcopy numbers used for standard curve constructions were shown in TableS2 in the supplemental material, along with the efficiency and sensitivityof the individual qPCR standard curves.

Statistical analyses. The coverage of each nifH clone library was cal-culated as C � [1 � (n1/N)] � 100, where n1 is the number of uniqueOTU and N is the total number of clones in a library (58). Indices of genediversity (Shannon-Wiener [H] and Simpson [D]) and evenness (J) werecalculated using the OTU data (50). Rarefaction analysis and two non-parametric richness estimators, the abundance-based coverage estimator(SACE) and the bias-corrected Chao1 (SChao1), were calculated usingDOTUR (53).

Community classification of the nifH-carrying microbial assemblageswas performed with Fast UniFrac environmental clustering and principalcoordinate analyses (PCoA) (59). Correlations between the nifH-harbor-ing microbial assemblages and environmental factors were determined bycanonical correspondence analysis (CCA) using the software Canoco(version 4.5; Microcomputer Power, Ithaca, NY) according to previouslydescribed procedures (50). Weighted NifH OTU and class data were usedto identify the most significant environmental factors that had the stron-gest influence on the community structure and spatial distribution of thenifH-harboring microbial assemblages in the nSCS.

Pearson correlation analyses (significance level � � 0.05) of the abun-dance of sediment 16S rRNA, nifH, nifD, and mcrA genes with environ-mental factors were performed with the statistics software MINITAB (re-lease 13.32; Minitab, State College, PA) as detailed previously (56, 57).

Nucleic acid sequence accession numbers. The nifH sequences deter-mined have been deposited in GenBank under accession numbersHQ223480 to HQ224498 and JQ412939 to JQ413029.

RESULTSSite description. Our sampling sites represent most of the typicalsedimentary environments of the nSCS (Fig. 1), including thePearl River estuary (station A3), coastal and offshore sites close toHainan Island (E501, E504, and E505, �200-m water depth) anddeep-water sites (1,000-m water depth) close to Taiwan Island(CF6), Luzon Island (E407), Dongsha Islands (CF11, CF14,and CF15), Xisha Trough (CF8 and E422), and Xisha Islands(E801). Methane hydrates have previously been discovered in theShenhu area southwest of Dongsha Islands (46). The XishaTrough, Jiulong methane reef area southwest of Taiwan Island,and Bijia’nan basin northwest of Luzon Island were previouslyidentified as gas hydrate prospective areas (43–45). Several of our

sampling sites were located in or near these previously identifiedor perspective methane hydrate reservoirs (see Fig. 1 for details).

Diversity of the sediment nifH genes. The two nifH gene clonelibraries (E407-I and E407-II) constructed from separate sedi-ment subcore samples of station E407 had similar OTU diver-sity based on rarefaction analysis (see Fig. S1 in the supplemen-tal material). Community classification using Fast UniFracenvironmental clustering (see Fig. S2 in the supplemental ma-terial) and PCoA (see Fig. S3 in the supplemental material)revealed that these two clone libraries were similar, indicatingboth the reproducibility of our experimental procedures andthe negligible within-site variability of the sediment nifH-har-boring microbial community at this station. Thus, these twonifH clone libraries were combined into a single library. Thehigh similarity of the nifH clone libraries from E407 distinctsubcore samples also suggests that the other sites have similarnegligible intrasite variability, making intersite comparisonsstatistically reliable and meaningful.

Of the 12 nifH clone libraries constructed for the nSCS sedi-ment samples, 1195 clones were identified to contain a valid nifHgene fragment, resulting in 1110 unique DNA sequences, 826unique protein sequences and 441 OTU. The values of librarycoverage (C) ranged from 43.6 to 66.3% (Table 2), which togetherwith rarefaction analysis (see Fig. S1 in the supplemental material)indicated a very high diversity of the nifH gene sequences in thesediments of the nSCS. The coastal site E501 had the lowest diver-sity and the offshore site E505 had the highest OTU diversity,based on all of the diversity indices (H, 1/D, and J). The SACE andSChao1 richness estimators are consistent to the above results forsite E505. However, these two estimators showed that the ex-pected OTU richness of site E501 were higher than some of theother sites (Table 2).

Phylogeny of the NifH protein sequences. The obtained 1,110distinct nifH gene sequences shared 40.2 to 99.8% identity withone another. Nearly half of these DNA sequences (49.5%) did nothave a match with known sequences from GenBank, except for theprimer regions. The sequences that did have matches were 64.8 to

TABLE 2 Biodiversity and predicted richness of the sediment nifH genesequences obtained from the sampling stations of the northern SouthChina Seaa

StationNo. ofclones

No. ofuniquegenesequences

No. ofOTU C (%) H 1/D J SACE SChao1

A3 91 86 52 59.3 5.17 27.67 0.91 186.83 112.55CF6 96 89 53 63.5 5.34 42.62 0.93 136.14 112.50CF8 80 79 47 60.0 5.17 35.51 0.93 101.30 109.00CF11 85 74 50 61.2 5.29 43.54 0.94 129.05 94.00CF14 89 86 56 51.7 5.42 43.51 0.93 232.32 206.50CF15 93 92 55 61.3 5.45 49.74 0.94 136.96 107.50E407 172 170 85 66.3 5.72 34.36 0.89 257.28 182.24E422 87 85 54 51.7 5.31 36.32 0.92 212.91 177.00E501 102 92 53 65.7 5.14 25.63 0.90 144.36 102.58E504 110 102 65 61.8 5.74 68.91 0.95 140.88 136.75E505 101 95 73 43.6 5.96 93.52 0.96 271.17 206.00E801 89 85 54 56.2 5.44 55.94 0.95 145.98 177.50a The OTU of the diazotrophic microbial NifH sequences were determined at a 0.05distance cutoff using the DOTUR program. The coverage (C), Shannon-Weiner (H),Simpson (D), and evenness (J) indices and the SACE and SChao1 richness estimators werecalculated using the OTU data.

Dang et al.

124 aem.asm.org Applied and Environmental Microbiology

on March 18, 2021 by guest

http://aem.asm

.org/D

ownloaded from

98.0% identical to the closest-match GenBank nifH sequences.However, all of the corresponding protein sequences deducedfrom the 1,110 nifH gene sequences showed various degrees ofidentity (42.1 to 100.0%) with known GenBank NifH sequences.The nSCS sediment NifH sequences were highly variable, and theyshared 25.2 to 100.0% sequence identity with one another. Nu-merous sequences had quite high identities (�90%) with NifHsequences from bacterial or archaeal isolates, such as those affili-ated with Alphaproteobacteria (Bradyrhizobium), Betaproteobacte-ria (Azoarcus), Gammaproteobacteria (Allochromatium, Azotobac-ter, Pseudomonas, Teredinibacter, and Vibrio), Deltaproteobacteria(Desulfatibacillum, Desulfobacterium, Desulfovibrio, Desulfuriv-ibrio, Desulfuromonas, Geobacter, and Pelobacter), Chlorobi (Chlo-robaculum, Chloroherpeton, and Prosthecochloris), Firmicutes(Clostridium, Desulfitobacterium, Desulfotomaculum, Paenibacil-lus, and Turicibacter), and methanogenic Euryarchaeota (Metha-nococcoides, Methanosarcina, Methanosphaerula, and “CandidatusMethanoregula”). Some sequences shared moderate identities(�80%) with NifH sequences from other bacterial or archaealisolates, such as those affiliated with Bacteroidetes (“CandidatusAzobacteroides”), Verrucomicrobia (Coraliomargarita and Verru-comicrobiae), Chlorobi (Chlorobium), Deltaproteobacteria (Desul-fonatronospira), Firmicutes (Alkaliphilus, Butyrivibrio, Ethanoli-genens, Eubacterium, and Syntrophothermus), and methanogenicEuryarchaeota (Methanohalophilus and Methanothermobacter).Half of the nSCS sediment NifH sequences (50.5%) shared loweridentities (�70%) with their best-match GenBank NifH se-quences.

The constructed phylogenetic tree putatively revealed a total ofnine NifH classes (Fig. 2; also see Fig. S4 in the supplementalmaterial), including all of the previously known NifH clusters (15)and several previously unidentified NifH clusters. A new putativeclassification schema for the NifH clusters emerged from our cur-rent phylogenetic analysis.

A total of 144 nSCS NifH OTU were affiliated within class I(Fig. 2; also see Fig. S4 in the supplemental material), containingNifH sequences from both bacteria and archaea and alternativenitrogenase reductase sequences encoded by anfH and vnfH (12).Several NifH clusters were identified in this class, including previ-ously defined clusters I, II, III, and IIIx, part (“subcluster A”) ofcluster IV (15), and two previously unknown clusters with NifHsequences exclusively from nSCS sediments (Fig. 2; also see Fig.S4 in the supplemental material). Our class I sequences wererelated to NifH sequences from diverse N2-fixing bacterial andarchaeal isolates, indicating that these nifH-harboring micro-organisms might be functional diazotrophs. The relatedGenBank environmental sequences were obtained from soils, rhizo-spheres of rice, soybean, mangrove, salt marsh, and seagrass, deep-seamethane seep sediments, and hydrothermal vent environments. Wealso obtained one sequence (E422-68) that was affiliated with theANME NifH cluster (“cluster IIIx”) previously identified exclusivelyfrom deep-sea methane seep sediments (14–16).

A total of 115 nSCS NifH OTU were affiliated within class II(Fig. 2; also see Fig. S4 in the supplemental material), which wasoriginally classified as part (“subcluster B”) of NifH “cluster IV” inprevious literature (9). Our NifH sequences, along with a few se-quences from methane-hydrate-bearing deep-sea sediments ofthe Okhotsk Sea and the deep-sea hydrothermal vent environ-ment of Juan de Fuca Ridge (12, 15), constituted the majority ofthe class II NifH sequences. The separation of these environmen-

tal sequences from the sequences obtained from bacterial and ar-chaeal isolates indicated the novelty of these marine NifH se-quences (see Fig. S4 in the supplemental material).

A total of 99 nSCS NifH OTU were affiliated within class III(Fig. 2; also see Fig. S4 in the supplemental material), which wasdefined as “cluster V” in a previous study (15). Sequences in thisclass were exclusively detected from sediments of the Okhotsk Seaand nSCS (15), except for NifH-2 from the non-N2-fixing Metha-nocella paludicola SANAE (GenBank accession numberBAI60977) and one Lake Vallentunasjon sequence that might bemistakenly classified as a chlorophyllide reductase-like sequencein the original publication (60).

Seventeen nSCS NifH OTU were affiliated within class IV (Fig.2; also see Fig. S4 in the supplemental material), which compriseda newly defined cluster in the present study. Sequence E407-II-40was related to some NifH sequences from Firmicutes (Dialister andSelenomonas), while all of the other nSCS sequences in this classwere distantly related to the NifH paralogues of methanogenicEuryarchaeota (Methanoculleus, Methanoplanus, Methanoregula,Methanosphaerula, and Methanospirillum).

Eleven nSCS NifH OTU were affiliated within class V (Fig. 2;also see Fig. S4 in the supplemental material), which was classifiedas part (“subcluster A”) of NifH “cluster VI” in a previous study(15). Our environmental sequences in this class were relatedto NifH sequences of anaerobic methanogenic Euryarchaeota(Methanosarcina and Methanospirillum), Deltaproteobacteria(Desulfatibacillum and Desulfobacterium), and Firmicutes (Desul-fotomaculum). Within this class, strain Methanosarcina acetiv-orans C2A was a diazotrophic archaeon (61), and strain Desulfati-bacillum alkenivorans AK-01 might be capable of carrying out N2

fixation (62). Thus, the microorganisms that carried the class VNifH sequences may be functional diazotrophs.

Forty-three nSCS NifH OTU were affiliated within class VI(Fig. 2; also see Fig. S4 in the supplemental material), which wasclassified as part (“subcluster B”) of NifH “cluster VI” in a previ-ous study (15). Our NifH sequences, along with a few sequencesfrom the methane-hydrate-bearing deep-sea sediments of theOkhotsk Sea (15), constituted the majority of the class VI NifHsequences. This class also includes some NifH sequences fromFusobacteria isolates (Fusobacterium) (63).

Four nSCS NifH OTU were affiliated within classes VII, VIII,and IX, respectively (Fig. 2; also see Fig. S4 in the supplementalmaterial), forming three newly defined NifH clusters. The NifHsequence from Clostridium cellulolyticum H10 is the only knownsequence that is affiliated with one of these classes.

Spatial distribution of the nifH-harboring microbial assem-blages. Multivariate statistical analyses indicated that differentnSCS sediment environments contained distinct nifH-harboringmicrobial assemblages. Both Fast UniFrac all-environment P testsignificance (P � 0.000) and UniFrac significance (P � 0.001)statistics indicated a significant difference among the nSCS sedi-ment nifH-harboring microbial assemblages. The heterogeneousdistribution of the nifH-harboring microbial communities wasconfirmed via Fast UniFrac PCoA (Fig. 3), environmental cluster-ing (see Fig. S5 in the supplemental material), and CCA (Fig. 4).Environmental variables in the first two CCA dimensions (CCA1and CCA2) explained 26.0% of the total variance in the nifH-harboring microbial community composition and 28.0% of thecumulative variance of the nifH-harboring microbial community-environment relationship (Fig. 4a). Sediment temperature was

South China Sea Deep-Sea Sediment nifH Diversity

January 2013 Volume 79 Number 1 aem.asm.org 125

on March 18, 2021 by guest

http://aem.asm

.org/D

ownloaded from

FIG 2 Skeleton phylogenetic tree simplified from the formal phylogenetic tree constructed from aligned NifH sequences as shown in Fig. S4 in the supplementalmaterial.

Dang et al.

126 aem.asm.org Applied and Environmental Microbiology

on March 18, 2021 by guest

http://aem.asm

.org/D

ownloaded from

identified as the only significant environmental factor (P � 0.008;1,000 Monte Carlo permutations) that contributed the most tothe heterogeneous community structure and spatial distributionof the nSCS sediment nifH-harboring microbial assemblages. Thisenvironmental factor alone provided 12.3% of the total CCA ex-planatory power.

Correlation of the nSCS sediment nifH-harboring microbialassemblages with environmental variables was further analyzedwith CCA using the NifH class data. Environmental variables inthe first two CCA dimensions (CCA1 and CCA2) explained 68.1%of the total variance in the sediment nifH-harboring microbialcommunity composition and 69.7% of the cumulative variance ofthe nifH-harboring microbial community-environment relation-

ship (Fig. 4b). Water depth (P � 0.035), sediment pore waterPO4

3� concentration (P � 0.051), and sediment temperature(P � 0.071) were identified as the only significant environmentalfactors, providing 11.5, 15.4, and 19.2% of the total CCA explan-atory power, respectively. All the other environmental factors an-alyzed were not found by the CCA to be significant (P 0.100) intheir respective contributions to the heterogeneous spatial distri-bution of the sediment nifH-harboring microbial assemblages inthe nSCS. The CCA showed that the sediment microorganismsthat harbored NifH sequences from distinct classes (Fig. 2; also seeFig. S4 in the supplemental material) responded differently to theprevalent in situ environmental factors (Fig. 4b). However, thetwo putatively functional diazotrophic groups that harbored theclass I and class V NifH sequences might have a similar environ-mental requirement that was different from the other nifH-har-boring microbial groups (Fig. 4b).

Abundance of nifH-harboring microorganisms. Meltingcurve analyses of the amplified genes (16S rRNA, nifH, nifD, andmcrA) confirmed that the fluorescence signals were obtained fromspecific PCR products of our qPCR quantifications. Standardcurves generated using plasmids containing cloned target genefragments to relate the threshold cycle (CT) to the gene copy num-ber revealed linearity (R2 � 0.981) over several orders of magni-tude of the standard plasmid DNA concentrations (see Table S2 inthe supplemental material). The obtained high correlation coeffi-cients and similar slopes indicated high primer hybridization andextension efficiencies (see Table S2 in the supplemental material),making comparison of the different genes’ abundances reliable.

The qPCR results showed heterogeneous distributions of thesediment bacterial and archaeal 16S rRNA gene abundances in thenSCS, with the shallow water sites usually harboring higher geneabundances for both bacteria and archaea (Table 3). In addition,the determined copy numbers of the bacterial 16S rRNA geneswere much higher than those of the archaeal 16S rRNA genes(averaging 23.5:1).

FIG 3 Ordination diagram of the Fast UniFrac weighted and normalizedPCoA analysis of the nSCS sediment nifH-harboring microbial assemblages asrevealed by using the NifH protein sequence data. Shown is the plot of the firsttwo principal coordinate axes (P1 and P2) for PCoA and the distributions ofnifH-harboring microbial assemblages (designated with the sampling stationnames) in response to these axes.

FIG 4 CCA ordination plots for the first two principal dimensions of the relationship between the in situ environmental parameters of the nSCS and thedistribution of the sediment nifH-harboring microbial assemblages as analyzed by using data of the NifH OTU (a) and the NifH classes (b). Correlations betweenenvironmental variables and CCA axes are represented by the length and angle of arrows (environmental factor vectors). Covarying variables, such as NH4

� andDIN (r � 0.9993), and OrgC and OrgN (r � 0.9680), were checked to minimize colinearity in the CCA analyses.

South China Sea Deep-Sea Sediment nifH Diversity

January 2013 Volume 79 Number 1 aem.asm.org 127

on March 18, 2021 by guest

http://aem.asm

.org/D

ownloaded from

The total nifH gene abundance also showed a distributionalheterogeneity in the nSCS sediments, where the shallow watersites usually harbored higher abundances (Table 3). The “clusterIIIx” nifH genes of the ANME diazotrophic group (15) were onlydetected in the sampling stations of methane hydrate or its pro-spective areas (CF6, CF11, CF14, CF15, E407, and E422), withquite low abundance (3.88 � 102 to 8.16 � 103 copies g of sedi-ment�1). Similarly, the nifD genes specific to the ANME di-azotrophs (17) were also detected only in these sampling stationswith low abundance (3.24 � 102 to 5.84 � 103 copies g of sedi-ment�1). Previous studies indicate that the ANME diazotrophsare exclusively associated with the ANME-2c subgroup (15–17).In the current study, the abundance of the ANME-2c subgroup-specific mcrA genes (i.e., mcrA subgroup c-d in Table 3) showedvery similar spatial distribution to the ANME diazotrophic nifHand nifD distributions in the nSCS sediments. The consistency ofthe qPCR quantification results from all of these three biomarkergenes indicated that the nSCS deep-sea sediments did harborANME diazotrophs, and they occurred only in methane hydrateor its prospective areas (Table 3).

The total archaeal mcrA gene abundance also showed a distri-butional heterogeneity: station E504 had the highest gene copynumber (1.83 � 106 copies g of sediment�1), and station CF8 hadthe lowest (2.75 � 105 copies g of sediment�1) (Table 3). Thesubgroup-specific mcrA genes that targeted the ANME subgroupsa-b, e, and f, respectively, also showed spatially heterogeneousdistributions in the sediments of the nSCS (Table 3).

The abundances of both sediment bacteria and archaea signif-icantly correlated positively with sediment temperature and neg-atively with water depth and sediment pore water dissolved oxy-gen content (DO) (see Table S3 in the supplemental material).The abundance of the nifH-harboring microbes significantly cor-related positively with sediment temperature, sediment pore wa-ter concentrations of NH4

� and DIN, and the sediment pore wa-ter N/P ratio and negatively with water depth, sediment inorganicphosphorus (IP), total phosphorus (TP), and organic matter(OM) contents. Moreover, most of the sedimentological parame-ters measured in the current study were also found to correlatewith the abundance of the nifH-harboring microbes (see Table S3in the supplemental material). The abundance of the sedimentANME diazotrophs significantly correlated positively with thesediment organic carbon, organic nitrogen, and organic phospho-

rus contents (OrgC, OrgN, and OrgP, respectively) in the nSCS, asdecoded by all three subgroup-specific target genes: ANME nifH,nifD, and subgroup c-d mcrA (see Table S3 in the supplementalmaterial).

The abundance of the sediment total archaeal methanogensand ANME methanotrophs detected by the total mcrA genes sig-nificantly correlated positively with sediment temperature andnegatively with water depth and sediment clay content in the nSCS(see Table S3 in the supplemental material). The abundance ofeach of the ANME subgroups detected by subgroup-specific mcrAgenes was correlated with distinct environmental factors, such aswater depth (positively) for mcrA subgroup a-b, sediment OrgC,OrgN, and OrgP (positively) for mcrA subgroup c-d, sedimenttemperature (positively) for mcrA subgroup e, and sedimentskewness (negatively) for mcrA subgroup f (see Table S3 in thesupplemental material).

DISCUSSIONDiversity and novelty of the nSCS nifH gene sequences. Diverseand novel nifH sequences were obtained in the current study (Ta-ble 2, Fig. 2; also see Fig. S1 and S4 in the supplemental material),expanding the NifH phylogeny from the previously defined sevenclusters (15) to the presently defined nine putative classes (Fig. 2;also see Fig. S4 in the supplemental material). Most of the newlydefined classes were established due to the discovery of novel nifHsequences from the sediments of the nSCS. Our new NifH classi-fication schema is more accurate than the previously defined ones(15). For example, the division of the NifH sequences of the pre-viously defined “cluster IV” and their inclusion in two distinctclasses (class I and class II) separately in our new NifH classifica-tion schema are consistent with the fact that the NifH sequencesfrom the previous “cluster IV” are polyphyletic (9, 15).

The great diversity and novelty of the nifH sequences obtainedfrom the nSCS marine sediments are highly unexpected as envi-ronmental nifH genes have been investigated for several decades(10). Technically, the success of our detection of a variety of novelnifH sequences and classes can be attributed to the use of the newnifH PCR primers designed by Mehta et al. (12). The unique en-vironment of nSCS sediments and the previous neglect of sedi-ment diazotrophic microbial communities may also have contrib-uted to the finding of novel nifH sequences and classes (Fig. 2 andsee Fig. S4 in the supplemental material). However, it should be

TABLE 3 Abundance of 16S rRNA, nifH, nifD, and mcrA genes in sediments of the 12 sampling stations of the northern South China Sea

Samplingstation

Mean no. of target genes g of sediment�1 (SE)

Bacterial 16S rRNA Archaeal 16S rRNA Total nifH ANME nifH ANME nifD

A3 1.52 � 1011 (3.84 � 1010) 5.47 � 109 (4.29 � 108) 2.63 � 107 (1.25 � 106) NDa NDCF6 2.68 � 1010 (3.57 � 109) 2.42 � 109 (8.43 � 107) 2.04 � 106 (1.76 � 103) 1.07 � 103 (2.69 � 10) 1.42 � 103 (6.45 � 10)CF8 2.82 � 1010 (1.37 � 109) 1.65 � 109 (1.59 � 108) 8.50 � 105 (7.19 � 104) ND NDCF11 6.17 � 1010 (1.51 � 1010) 1.20 � 109 (8.42 � 107) 1.60 � 106 (5.46 � 104) 3.31 � 103 (1.15 � 102) 3.59 � 103 (5.79 � 10)CF14 4.93 � 1010 (2.30 � 109) 1.16 � 109 (3.07 � 107) 3.17 � 106 (2.64 � 105) 4.59 � 103 (2.24 � 102) 3.89 � 103 (6.06 � 10)CF15 1.30 � 1011 (5.59 � 109) 4.95 � 109 (2.67 � 107) 1.35 � 106 (1.22 � 105) 8.16 � 103 (4.65 � 102) 5.84 � 103 (1.69 � 102)E407 2.80 � 1010 (2.49 � 109) 1.51 � 109 (4.53 � 107) 9.83 � 105 (3.58 � 104) 5.94 � 102 (3.97 � 10) 3.24 � 102 (3.03 � 10)E422 3.05 � 1010 (2.31 � 109) 1.05 � 109 (7.62 � 107) 1.64 � 106 (2.95 � 104) 3.88 � 102 (3.29 � 10) 1.55 � 103 (6.88 � 10)E501 9.39 � 1010 (1.01 � 1010) 6.03 � 109 (4.42 � 108) 5.65 � 107 (3.79 � 106) ND NDE504 1.44 � 1011 (5.15 � 1010) 1.73 � 1010 (9.52 � 108) 1.20 � 107 (1.09 � 106) ND NDE505 6.07 � 1010 (5.33 � 109) 5.48 � 109 (2.34 � 108) 4.34 � 106 (3.40 � 104) ND NDE801 3.13 � 1010 (2.19 � 109) 1.32 � 109 (2.34 � 107) 3.18 � 106 (3.97 � 104) ND ND

(Continued on next page)

Dang et al.

128 aem.asm.org Applied and Environmental Microbiology

on March 18, 2021 by guest

http://aem.asm

.org/D

ownloaded from

borne in mind that probably not all of the detected nifH sequencescame from active diazotrophic microorganisms. For example,most of the NifH sequences from the originally defined “clusterIV” are not well characterized, and some are found not to beinvolved in N2 fixation (9, 17, 64). Although it is currently un-known whether the microorganisms that carry the novel nifH se-quences from most of our newly defined classes are functional foractive N2 fixation in natural environments, our study extends thediversity and evolutionary complexity of the nifH gene sequences.Recent genome sequencing of various bacterial and archaeal iso-lates provides further evidence of the diversity and novelty of thenifH sequences in the microbial world (see Fig. S4 in the supple-mental material for some of the nifH sequences from microbialgenomes). Furthermore, some nifH genes are found to be locatedon plasmids (65). Rapid evolution and high frequency of horizon-tal gene transfer of the nifH genes, even at the interdomain levelbetween bacteria and archaea (9), may favor diversity and noveltyof the nifH sequences, especially in the sediment environment,where intimate syntrophic interactions between bacteria and ar-chaea, such as the consortia formed between sulfate-reducing bac-teria and ANME (14, 16), may facilitate microbial gene transfer.The deep ocean is a source of abundant nitrate (66). Sediment N2

fixation by diverse nifH-harboring bacteria and archaea may con-tribute significantly to this large nutrient reservoir.

Environmental influence on the sediment nifH-harboringmicrobial assemblages. Abundant bacteria and archaea were de-tected in the marine sediments of the nSCS, with bacterial 16SrRNA gene copies being predominant over those of archaea (Ta-ble 3). Using one of the most efficient and successful methods(FastPrep) for sediment DNA extraction (67), determined 16SrRNA gene copy numbers of bacteria were much higher thanthose of archaea (averaging 23.5:1). Even with consideration ofvarious 16S rRNA gene copy numbers in bacterial and archaealgenomes and the potential for unequal DNA extraction andprimer specificity for the molecular detection (68, 69), the largeratio suggests that bacteria dominate quantitatively over archaeain the surface sediments of the nSCS.

The correlation of the total bacterial and archaeal abundanceswith sediment temperature (positively), sediment pore water DO(negatively), and water depth (negatively) (see Table S3 in thesupplemental material) suggests that the sediment bacterial andarchaeal abundances decrease from coastal to deep-sea environ-

ments, likely controlled by sediment in situ temperature and porewater DO. These results also indicate that anaerobic bacteria andarchaea may be the major constituents of the microbiota in thesurface sediments (down to a 5-cm depth) of the nSCS.

The correlation of the total nifH-harboring microbial abun-dance with sediment temperature (positively), water depth (neg-atively), and most of the sedimentological parameters (see TableS3 in the supplemental material) indicates that the surface sedi-ment nifH-harboring microbial abundance decreases from coastalto deep-sea environments, likely controlled by in situ sedimenttemperature and sedimentological conditions. CCA also indicatedthat sediment temperature was an important environmentalfactor influencing the community structure and spatial distribu-tion of the sediment nifH-harboring microbial assemblages in thenSCS (Fig. 4). It has been found previously that temperature is themost important environmental factor influencing the spatial dis-tribution of seawater N2-fixing cyanobacteria in the oceans (70).Thus, temperature may be a universally important environmentalfactor controlling the distribution of the nifH-harboring micro-bial community in both water columns and sediments in marineenvironments.

The abundance of the sediment total nifH genes also correlatedpositively with the sediment pore water N/P ratio and concentra-tions of NH4

� and DIN (see Table S3 in the supplemental mate-rial), indicating that the sediment nifH-harboring microorgan-isms might be active in N2 fixation and NH4

� production in thenSCS. We also found that the abundance of the sediment totalnifH genes correlated negatively with sediment inorganic P (IP),total P (TP), and organic matter (OM) contents (see Table S3 inthe supplemental material). This might indicate that the sedimentnifH-harboring microbial population actively consumed sedi-ment phosphorus and organic matter. However, this explanationneeds to be viewed with caution. Most of the sediment P might betightly absorbed or bound to sediment particles or minerals andthus might not be the soluble reactive phosphorus for microbialutilization. Similarly, most deep-sea sediment organic mattermight be old and recalcitrant to microbial utilization. Therefore,whether the negative correlations of the sediment nifH-harboringmicroorganism abundance with the sediment inorganic P, total P,and organic matter contents might indicate a cause-effect rela-tionship needs to be further investigated.

CCA identified sediment pore water PO43� concentration as a

TABLE 3 (Continued)

Mean no. of target genes g of sediment�1 (SE)

Total mcrA Subgroup a-b mcrA Subgroup c-d mcrA Subgroup e mcrA Subgroup f mcrA

6.80 � 105 4.18 � 104 ND ND 4.45 � 104 (2.91 � 102) 2.36 � 103 (1.61 � 102)4.58 � 105 1.02 � 104 1.69 � 103 (1.19 � 102) 1.62 � 103 (3.02 � 10) ND 1.86 � 103 (1.79 � 102)2.75 � 105 6.92 � 103 2.56 � 10 (2.22) ND ND 4.81 � 102 (1.64 � 10)3.05 � 105 1.71 � 104 1.26 � 102 (5.91) 4.27 � 103 (3.98 � 102) 2.55 � 103 (2.09 � 102) 3.53 � 102 (1.91 � 10)5.78 � 105 4.44 � 104 2.40 � 102 (2.26 � 10) 3.87 � 103 (2.75 � 10) 1.37 � 103 (7.81) 6.62 � 102 (5.66 � 10)3.30 � 105 1.90 � 104 4.39 � 102 (1.59 � 10) 5.58 � 103 (1.39 � 102) 3.23 � 103 (1.89 � 10) 1.79 � 103 (1.63 � 102)5.12 � 105 5.81 � 103 3.87 � 102 (1.10 � 10) 3.67 � 102 (9.45) 2.32 � 103 (2.99 � 10) 3.36 � 103 (3.71 � 10)3.50 � 105 2.52 � 104 7.84 � 102 (6.82 � 10) 4.97 � 102 (2.18 � 10) 2.09 � 103 (1.58 � 102) 4.47 � 102 (3.95 � 10)9.57 � 105 4.62 � 104 ND ND 4.78 � 103 (1.36 � 102) ND1.83 � 106 9.04 � 104 1.12 � 102 (6.24) ND 3.97 � 103 (3.89 � 102) ND1.61 � 106 3.67 � 104 1.43 � 102 (1.06 � 10) ND 3.48 � 103 (9.64 � 10) ND7.41 � 105 5.33 � 104 ND ND ND NDa ND, not detectable.

South China Sea Deep-Sea Sediment nifH Diversity

January 2013 Volume 79 Number 1 aem.asm.org 129

on March 18, 2021 by guest

http://aem.asm

.org/D

ownloaded from

key environmental factor influencing the community structureand spatial distribution of the nifH-harboring microorganisms inthe sediments of the nSCS (Fig. 4b). This is meaningful since porewater PO4

3� is the major form of soluble reactive phosphorusutilizable by microorganisms. The nSCS has experienced somedramatic change during the last several decades, mainly caused byintensified anthropogenic activities. Recent studies indicated thatthe bioproductivity of nSCS might be limited by P in estuarine andcoastal areas (71, 72). Our data showed that the two shallowestsampling sites, A3 and E501, indeed had quite low sediment porewater PO4

3� concentrations (Table 1). It is interesting to discoverthat the availability of sediment pore water PO4

3� may have animportant influence on the sediment nifH-harboring microbialcommunity structure and distribution and thus potentially on thesediment nitrogen-fixing activity.

Deep-sea methane seep-specific ANME-2c diazotrophic ar-chaea. Methane seeps, mainly formed in gas hydrate-bearing sed-iments of marginal seas, sustain significant chemosynthetic eco-systems (73). The characteristic high biomass and productivity ofthese ecosystems, driven by rapid C and S microbial transforma-tions, may concomitantly require a large supply of N nutrients(15, 56). Recent investigations indicate that N2-fixing microor-ganisms, especially archaea in the ANME-2c subgroup, may play acritical role in supplying newly fixed nitrogen to these ecosystems(14–17). These studies also indicate that the distribution of theANME-2c subgroup diazotrophs may be methane seep specific,seemingly absent in other nonseep marine environments (15–17).The results of the present study support this hypothesis.

Our quantification results of universal and subgroup-specificqPCR determinations of the mcrA genes indicated that archaealmethanogens were present in all of the sediment environments ofthe nSCS, while distribution of the specific ANME subgroups var-ied and might be controlled by distinct environmental factors (Ta-ble 3; also see Table S3 in the supplemental material). These resultsalso indicated that only the abundance of subgroup c-d mcrAgenes correlated with the abundances of the “cluster IIIx” nifHand nifD genes (Table 3), suggesting that only the ANME-2carchaea and not the archaea from other ANME subgroups werediazotrophs. This finding agrees with previous studies (14–17).Correlation analyses using the mcrA subgroup-specific qPCR dataindicated that the ANME-2c diazotrophs had a distribution pat-tern and environment relationship distinct from those of the othernondiazotrophic ANME subgroups (see Table S3 in the supple-mental material). ANME-2c diazotroph abundance significantlycorrelated with sediment OrgC, OrgN, and OrgP for all three“cluster IIIx”-specific target genes: ANME nifH, nifD, and sub-group c-d mcrA (see Table S3 in the supplemental material). Al-though ANME activity is usually thought of as a chemolithoau-totrophic process, certain organic substrates may provide optimalgrowth and performances of anaerobic methane oxidation and N2

fixation for the syntrophic consortia of ANME archaea and sul-fate-reducing bacteria (14, 16, 74).

Quantifications via qPCR targeting the “cluster IIIx”-specificnifH, nifD, and mcrA genes confirmed the existence of theANME-2c diazotrophs in the marine sediments of the methanehydrate-bearing or prospective areas in the nSCS (Table 3). Al-though only one ANME-2c diazotroph-related nifH sequence(E422-68, GenBank accession no. HQ224110) was obtained in thepresent study (see Fig. S4 in the supplemental material), the rareoccurrence of the ANME nifH sequences in our clone libraries is

not a surprise. The ANME-2c diazotrophs might exist with verylow abundance in the marine sediments and their correspondingnifH sequences might be diluted out in the clone libraries due tothe overwhelmingly high diversity and abundance of the othernifH sequences in the nSCS sediments (Table 2, Fig. 2; also see Fig.S1 and S4 in the supplemental material). The nSCS methane hy-drates are buried very deeply in the sediments (150 m below theseafloor surface), and the nSCS methane seeps are characterizedby micro gas venting with conduit/channel diameters of only 200to 600 �m (46, 75). The deep burial of methane hydrates and thelow level of methane gas venting make the intensity of the meth-ane supply very low and also the methane seeps as localized fea-tures in the surface sediments of the nSCS. Although some of oursediment sampling sites were in or near the methane hydrate-bearing or prospective areas, they were likely not located exactlyon the main venting channels due to the difficulty of on-boarddeep-sea sediment sampling. A low supply of methane might sus-tain low ANME diazotrophic populations, as detected by ourqPCR measurements (Table 3). Our current data indicated thatthe distribution of the ANME-2c diazotrophs was associated withthe presence of methane seeps, possibly caused by the dependenceof ANME archaea on the methane supply for their ecophysiologi-cal activities. Our current data, together with the results fromother recent studies (14–17), support the hypothesis thatANME-2c diazotrophs reside exclusively in the deep-sea methaneseep sediments and that their presence has diagnostic value for thediscovery of deep-sea methane hydrate reservoirs.

ACKNOWLEDGMENTS

This study was supported by China NSFC grants 91028011 (H.D.) and41076091 (H.D. and M.G.K.), National Key Basic Research Program ofChina grants 2013CB955700 (H.D.) and 2007CB411702 (X.L.), Funda-mental Research Funds for the Central Universities of China grant09CX05005A (H.D.), and U.S. NSF grants 0541797 and 0948202(M.G.K.).

The sediment samples used in this study were collected during the2007 South China Sea Open Cruise by R/V Shiyan 3, South China SeaInstitute of Oceanology, CAS. We thank two anonymous reviewers fortheir constructive comments and Jian Ren, Shuai Wang, LongshengCheng, Songbing Yan, Zihan Chen, Wei Wei, and Guanhang Liu for theirassistance in the project.

REFERENCES1. Zehr JP, Jenkins BD, Short SM, Steward GF. 2003. Nitrogenase gene

diversity and microbial community structure: a cross-system comparison.Environ. Microbiol. 5:539 –554.

2. Ward BB, Capone DG, Zehr JP. 2007. What’s new in the nitrogen cycle?Oceanography 20:101–109.

3. Codispoti LA. 2007. An oceanic fixed nitrogen sink exceeding 400 Tg Na�1 versus the concept of homeostasis in the fixed-nitrogen inventory.Biogeosciences 4:233–253.

4. Fulweiler RW. 2009. Fantastic fixers. Science 326:377–378.5. Bonnet S, Biegala IC, Dutrieux P, Slemons LO, Capone DG. 2009.

Nitrogen fixation in the western equatorial Pacific: rates, diazotrophiccyanobacterial size class distribution, and biogeochemical significance.Global Biogeochem. Cycles 23:GB3012.

6. Grosse J, Bombar D, Doan HN, Nguyen LN, Voss M. 2010. The MekongRiver plume fuels nitrogen fixation and determines phytoplankton speciesdistribution in the South China Sea during low- and high-discharge sea-son. Limnol. Oceanogr. 55:1668 –1680.

7. Moisander PH, Beinart RA, Hewson I, White AE, Johnson KS, CarlsonCA, Montoya JP, Zehr JP. 2010. Unicellular cyanobacterial distributionsbroaden the oceanic N2 fixation domain. Science 327:1512–1514.

8. Fernandez C, Farías L, Ulloa O. 2011. Nitrogen fixation in denitrifiedmarine waters. PLoS One 6:e20539. doi:10.1371/journal.pone.0020539.

Dang et al.

130 aem.asm.org Applied and Environmental Microbiology

on March 18, 2021 by guest

http://aem.asm

.org/D

ownloaded from

9. Raymond J, Siefert JL, Staples CR, Blankenship RE. 2004. The naturalhistory of nitrogen fixation. Mol. Biol. Evol. 21:541–554.

10. Zehr JP, Capone DG. 1996. Problems and promises of assaying the ge-netic potential for nitrogen fixation in the marine environment. Microb.Ecol. 32:263–281.

11. Lovell CR. 2002. Plant-microbe interactions in the marine environment,p 2539 –2554. In Bitton G (ed), Encyclopedia of environmental microbi-ology, vol 5. Wiley, New York, NY.

12. Mehta MP, Butterfield DA, Baross JA. 2003. Phylogenetic diversity ofnitrogenase (nifH) genes in deep-sea and hydrothermal vent environ-ments of the Juan de Fuca Ridge. Appl. Environ. Microbiol. 69:960 –970.

13. Mehta MP, Huber JA, Baross JA. 2005. Incidence of novel and poten-tially archaeal nitrogenase genes in the deep Northeast Pacific Ocean. En-viron. Microbiol. 7:1525–1534.

14. Pernthaler A, Dekas AE, Brown CT, Goffredi SK, Embaye T, OrphanVJ. 2008. Diverse syntrophic partnerships from deep-sea methane ventsrevealed by direct cell capture and metagenomics. Proc. Natl. Acad. Sci.U. S. A. 105:7052–7057.

15. Dang HY, Luan XW, Zhao JY, Li J. 2009. Diverse and novel nifH andnifH-like gene sequences in the deep-sea methane seep sediments of theOkhotsk Sea. Appl. Environ. Microbiol. 75:2238 –2245.

16. Dekas AE, Poretsky RS, Orphan VJ. 2009. Deep-sea archaea fix and sharenitrogen in methane-consuming microbial consortia. Science 326:422–426.

17. Miyazaki J, Higa R, Toki T, Ashi J, Tsunogai U, Nunoura T, Imachi H,Takai K. 2009. Molecular characterization of potential nitrogen fixationby anaerobic methane-oxidizing archaea in the methane seep sediments atthe number 8 Kumano Knoll in the Kumano Basin, offshore of Japan.Appl. Environ. Microbiol. 75:7153–7162.

18. Gaby JC, Buckley DH. 2011. A global census of nitrogenase diversity.Environ. Microbiol. 13:1790 –1799.

19. Whitman WB, Coleman DC, Wiebe WJ. 1998. Prokaryotes: the unseenmajority. Proc. Natl. Acad. Sci. U. S. A. 95:6578 – 6583.

20. Zehr JP, Kudela RM. 2011. Nitrogen cycle of the open ocean: from genesto ecosystems. Annu. Rev. Mar. Sci. 3:197–225.

21. Zehr JP, Mellon M, Braun S, Litaker W. 1995. Diversity of heterotrophicnitrogen fixation genes in a marine cyanobacterial mat. Appl. Environ.Microbiol. 61:2527–2532.

22. Burns JA, Zehr JP, Capone DG. 2002. Nitrogen-fixing phylotypes ofChesapeake Bay and Neuse River estuary sediments. Microb. Ecol. 44:336 –343.

23. Flores-Mireles AL, Winans SC, Holguin G. 2007. Molecular character-ization of diazotrophic and denitrifying bacteria associated with man-grove roots. Appl. Environ. Microbiol. 73:7308 –7321.

24. Charpy L, Palinska KA, Casareto B, Langlade MJ, Suzuki Y, Abed RM,Golubic S. 2010. Dinitrogen-fixing cyanobacteria in microbial mats oftwo shallow coral reef ecosystems. Microb. Ecol. 59:174 –186.

25. Gamble MD, Bagwell CE, Larocque J, Bergholz PW, Lovell CR. 2010.Seasonal variability of diazotroph assemblages associated with the rhizo-sphere of the salt marsh cordgrass, Spartina alterniflora. Microb. Ecol.59:253–265.

26. Severin I, Acinas SG, Stal LJ. 2010. Diversity of nitrogen-fixing bacteriain cyanobacterial mats. FEMS Microbiol. Ecol. 73:514 –525.

27. Walsh JJ. 1991. Importance of continental margins in the marine biogeo-chemical cycling of carbon and nitrogen. Nature 350:53–55.

28. Bauer JE, Druffel ERM. 1998. Ocean margins as a significant source oforganic matter to the deep open ocean. Nature 392:482– 485.

29. Subramaniam A, Yager PL, Carpenter EJ, Mahaffey C, Björkman K,Cooley S, Kustka AB, Montoya JP, Sañudo-Wilhelmy SA, Shipe R,Capone DG. 2008. Amazon River enhances diazotrophy and carbon se-questration in the tropical North Atlantic Ocean. Proc. Natl. Acad. Sci.U. S. A. 105:10460 –10465.

30. Capone DG, Burns JA, Michaels AF, Montoya JP, Subramaniam A,Carpenter EJ. 2005. Nitrogen fixation by Trichodesmium spp.: an impor-tant source of new nitrogen to the tropical and sub-tropical North AtlanticOcean. Global Biogeochem. Cycles 19:GB2024.

31. Sohm JA, Webb EA, Capone DG. 2011. Emerging patterns of marinenitrogen fixation. Nat. Rev. Microbiol. 9:499 –508.

32. Knittel K, Boetius A. 2009. Anaerobic oxidation of methane: progresswith an unknown process. Annu. Rev. Microbiol. 63:311–334.

33. Hester KC, Brewer PG. 2009. Clathrate hydrates in nature. Annu. Rev.Mar. Sci. 1:303–327.

34. Moisander PH, Beinart RA, Voss M, Zehr JP. 2008. Diversity and

abundance of diazotrophic microorganisms in the South China Sea dur-ing intermonsoon. ISME J. 2:954 –967.

35. Wu J, Chung SW, Wen LS, Liu KK, Chen YL, Chen H, Karl DM. 2003.Dissolved inorganic phosphorus, dissolved iron, and Trichodesmium inthe oligotrophic South China Sea. Global Biogeochem. Cycles 17:1008.

36. Yin KD, Qian Wu P-YMCS, Chen JC, Huang LM, Song XY, Jian WJ.2001. Shift from P to N limitation of phytoplankton growth across thePearl River estuarine plume during summer. Mar. Ecol. Prog. Ser.221:17–28.

37. Chen YL, Chen H, Karl DM, Takahashi M. 2004. Nitrogen modulatesphytoplankton growth in spring in the South China Sea. Cont. Shelf Res.24:527–541.

38. Zhang Y, Zhao ZH, Sun J, Jiao NZ. 2011. Diversity and distribution ofdiazotrophic communities in the South China Sea deep basin with me-soscale cyclonic eddy perturbations. FEMS Microbiol. Ecol. 78:417– 427.

39. Chou W, Chen YL, Sheu DD, Shih Y, Han C, Cho CL, Tseng C, YangY. 2006. Estimated net community production during the summertime atthe SEATS time-series study site, northern South China Sea: implicationsfor nitrogen fixation. Geophys. Res. Lett. 33:L22610.

40. Voss M, Bombar D, Natalie L, Dippner JW. 2006. Riverine influence onnitrogen fixation in the upwelling region off Vietnam, South China Sea.Geophys. Res. Lett. 33:L07604.

41. Chen YL, Chen H, Tuo S, Ohki K. 2008. Seasonal dynamics of newproduction from Trichodesmium N2 fixation and nitrate uptake in theupstream Kuroshio and South China Sea basin. Limnol. Oceanogr. 53:1705–1721.

42. Bombar D, Moisander PH, Dippner JW, Foster RA, Voss M, Karfeld B,Zehr JP. 2011. Distribution of diazotrophic microorganisms and nifHgene expression in the Mekong River plume during intermonsoon. Mar.Ecol. Prog. Ser. 424:39 –52.

43. Zhu YH, Wu BH, Lu ZQ. 2001. The prospecting for gas hydrate offshoreChina. Mineral Deposits 20:174 –180. (In Chinese with English abstract.)

44. Chen DF, Huang YY, Yuan XL, Cathles LM III. 2005. Seep carbonatesand preserved methane oxidizing archaea and sulfate reducing bacteriafossils suggest recent gas venting on the seafloor in the northeastern SouthChina Sea. Mar. Petrol. Geol. 22:613– 621.

45. Han XQ, Suess E, Huang YY, Wu NY, Bohrmann G, Su X, EisenhauerA, Rehder G, Fang YX. 2008. Jiulong methane reef: microbial mediationof seep carbonates in the South China Sea. Mar. Geol. 249:243–256.

46. Wu NY, Zhang HQ, Yang SX, Zhang GX, Liang JQ, Lu JA, Su X,Schultheiss P, Holland M, Zhu YH. 2011. Gas hydrate system of Shenhuarea, northern South China Sea: geochemical results. J. Geol. Res. 2011:370298.

47. Zhou MY, Chen XL, Zhao HL, Dang HY, Luan XW, Zhang XY, He HL,Zhou BC, Zhang YZ. 2009. Diversity of both the cultivable protease-producing bacteria and their extracellular proteases in the sediments ofthe South China Sea. Microb. Ecol. 58:582–590.

48. Danovaro R. 2009. Methods for the study of deep-sea sediments, theirfunctioning, and biodiversity. CRC Press, Boca Raton, FL.

49. Grasshoff K, Kremling K, Ehrhardt M. 1999. Methods of seawater anal-ysis. Wiley-Verlag Chemie, Weinheim, Germany.

50. Dang HY, Zhang XX, Sun J, Li TG, Zhang ZN, Yang GP. 2008. Diversityand spatial distribution of sediment ammonia-oxidizing Crenarchaeota inresponse to estuarine and environmental gradients in the ChangjiangEstuary and East China Sea. Microbiology 154:2084 –2095.

51. Dang HY, Li TG, Chen MN, Huang GQ. 2008. Cross-ocean distributionof Rhodobacterales bacteria as primary surface colonizers in temperatecoastal marine waters. Appl. Environ. Microbiol. 74:52– 60.

52. Altschul S, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W,Lipman DJ. 1997. Gapped BLAST and PSI-BLAST: a new generation ofprotein database search programs. Nucleic Acids Res. 25:3389 –3402.

53. Schloss PD, Handelsman J. 2005. Introducing DOTUR, a computerprogram for defining operational taxonomic units and estimating speciesrichness. Appl. Environ. Microbiol. 71:1501–1506.

54. Larkin MA, Blackshields G, Brown NP, Chenna R, McGettigan PA,McWilliam H, Valentin F, Wallace IM, Wilm A, Lopez R, ThompsonJD, Gibson TJ, Higgins DG. 2007. CLUSTAL W and CLUSTAL X version2.0. Bioinformatics 23:2947–2948.

55. Felsenstein J. 1989. PHYLIP: phylogeny inference package (version 3.2).Cladistics 5:164 –166.

56. Dang HY, Luan XW, Chen RP, Zhang XX, Guo LZ, Klotz MG. 2010.Diversity, abundance and distribution of amoA-encoding archaea in

South China Sea Deep-Sea Sediment nifH Diversity

January 2013 Volume 79 Number 1 aem.asm.org 131

on March 18, 2021 by guest

http://aem.asm

.org/D

ownloaded from

deep-sea methane seep sediments of the Okhotsk Sea. FEMS Microbiol.Ecol. 72:370 –385.

57. Dang HY, Chen RP, Wang L, Guo LZ, Chen PP, Tang ZW, Tian F, LiSZ, Klotz MG. 2010. Environmental factors shape sediment anammoxbacterial communities in hypernutrified Jiaozhou Bay, China. Appl. En-viron. Microbiol. 76:7036 –7047.

58. Mullins TD, Britschgi TB, Krest RL, Giovannoni SJ. 1995. Geneticcomparisons reveal the same unknown bacterial lineages in Atlantic andPacific bacterioplankton communities. Limnol. Oceanogr. 40:148 –158.

59. Hamady M, Lozupone C, Knight R. 2010. Fast UniFrac: facilitatinghigh-throughput phylogenetic analyses of microbial communities includ-ing analysis of pyrosequencing and PhyloChip data. ISME J. 4:17–27.

60. Eiler A, Beier S, Säwström C, Karlsson J, Bertilsson S. 2009. High ratioof bacteriochlorophyll biosynthesis genes to chlorophyll biosynthesisgenes in bacteria of humic lakes. Appl. Environ. Microbiol. 75:7221–7228.

61. Galagan JE, Nusbaum C, Roy A, Endrizzi MG, Macdonald P, FitzHughW, Calvo S, Engels R, Smirnov S, Atnoor D, Brown A, Allen N, NaylorJ, Stange-Thomann N, DeArellano K, Johnson R, Linton L, McEwan P,McKernan K, Talamas J, Tirrell A, Ye W, Zimmer A, Barber RD, CannI, Graham DE, Grahame DA, Guss AM, Hedderich R, Ingram-Smith C,Kuettner HC, Krzycki JA, Leigh JA, Li W, Liu J, Mukhopadhyay B,Reeve JN, Smith K, Springer TA, Umayam LA, White O, White RH,Conway DE Macario E, Ferry JG, Jarrell KF, Jing H, Macario AJ,Paulsen I, Pritchett M, Sowers KR, Swanson RV, Zinder SH, Lander E,Metcalf WW, Birren B. 2002. The genome of Methanosarcina acetivoransreveals extensive metabolic and physiological diversity. Genome Res. 12:532–542.

62. Callaghan AV, Morris BE, Pereira IA, McInerney MJ, Austin RN,Groves JT, Kukor JJ, Suflita JM, Young LY, Zylstra GJ, Wawrik B. 2012.The genome sequence of Desulfatibacillum alkenivorans AK-01: a blue-print for anaerobic alkane oxidation. Environ. Microbiol. 14:101–113.

63. Kapatral V, Anderson I, Ivanova N, Reznik G, Los T, Lykidis A,Bhattacharyya A, Bartman A, Gardner W, Grechkin G, Zhu L, VasievaO, Chu L, Kogan Y, Chaga O, Goltsman E, Bernal A, Larsen N, D’SouzaM, Walunas T, Pusch G, Haselkorn R, Fonstein M, Kyrpides N, Over-beek R. 2002. Genome sequence and analysis of the oral bacterium Fuso-bacterium nucleatum strain ATCC 25586. J. Bacteriol. 184:2005–2018.

64. Staples CR, Lahiri S, Raymond J, Von Herbulis L, Mukhophadhyay B,Blankenship RE. 2007. Expression and association of group IV nitroge-

nase NifD and NifH homologs in the non-nitrogen-fixing archaeonMethanocaldococcus jannaschii. J. Bacteriol. 189:7392–7398.

65. Heidelberg JF, Seshadri R, Haveman SA, Hemme CL, Paulsen IT,Kolonay JF, Eisen JA, Ward N, Methe B, Brinkac LM, Daugherty SC,Deboy RT, Dodson RJ, Durkin AS, Madupu R, Nelson WC, SullivanSA, Fouts D, Haft DH, Selengut J, Peterson JD, Davidsen TM, Zafar N,Zhou L, Radune D, Dimitrov G, Hance M, Tran K, Khouri H, Gill J,Utterback TR, Feldblyum TV, Wall JD, Voordouw G, Fraser CM. 2004.The genome sequence of the anaerobic, sulfate-reducing bacterium Desul-fovibrio vulgaris Hildenborough. Nat. Biotechnol. 22:554 –559.

66. Lam P, Jensen MM, Lavik G, McGinnis DF, Muller B, Schubert CJ,Amann R, Thamdrup B, Kuypers MM. 2007. Linking crenarchaeal andbacterial nitrification to anammox in the Black Sea. Proc. Natl. Acad. Sci.U. S. A. 104:7104 –7109.

67. Webster G, Newberry CJ, Fry JC, Weightman AJ. 2003. Assessment ofbacterial community structure in the deep sub-seafloor biosphere by 16SrDNA-based techniques: a cautionary tale. J. Microbiol. Methods 55:155–164.

68. Acinas SG, Marcelino LA, Klepac-Ceraj V, Polz MF. 2004. Divergenceand redundancy of 16S rRNA sequences in genomes with multiple rrnoperons. J. Bacteriol. 186:2629 –2635.

69. Lipp JS, Morono Y, Inagaki F, Hinrichs KU. 2008. Significant contri-bution of archaea to extant biomass in marine subsurface sediments. Na-ture 454:991–994.

70. Stal LJ. 2009. Is the distribution of nitrogen-fixing cyanobacteria in theoceans related to temperature? Environ. Microbiol. 11:1632–1645.

71. Xu J, Yin KD, He L, Yuan XC, Ho AYT, Harrison PJ. 2008. Phosphoruslimitation in the northern South China Sea during late summer: influenceof the Pearl River. Deep Sea Res. 55:1330 –1342.

72. Ning X, Lin C, Hao Q, Liu C, Le F, Shi J. 2009. Long term changes in theecosystem in the northern South China Sea during 1976 –2004. Biogeosci-ences 6:2227–2243.

73. Levin LA. 2005. Ecology of cold seep sediments: interactions of fauna withflow, chemistry and microbes. Oceanogr. Mar. Biol. Annu. Rev. 43:1– 46.

74. Stams AJ, Plugge CM. 2009. Electron transfer in syntrophic communitiesof anaerobic bacteria and archaea. Nat. Rev. Microbiol. 7:568 –577.

75. Chen Z, Yan W, Chen MH, Wang SH, Lu J, Zhang F, Xiang R, Xiao SB,Yan B, Gu SC. 2006. Discovery of seep carbonate nodules as new evidencefor gas venting on the northern continental slope of South China Sea.Chin. Sci. Bull. 51:1228 –1237.

Dang et al.

132 aem.asm.org Applied and Environmental Microbiology

on March 18, 2021 by guest

http://aem.asm

.org/D

ownloaded from