Arch Microbiol (2013) 195:853–859 DOI 10.1007/s00203-013-0937-z

Short Communication

Diazotrophic diversity in the Caribbean coral, Montastraea cavernosa Nathan D. Olson · Michael P. Lesser 

Received: 9 April 2013 / Revised: 5 September 2013 / Accepted: 24 October 2013 / Published online: 12 November 2013 © Springer-Verlag Berlin Heidelberg 2013

Abstract  Previous research on the Caribbean coral Montastraea cavernosa reported the presence of cyanobacterial endosymbionts and nitrogen fixation in orange, but not brown, colonies. We compared the diversity of nifH gene sequences between these two color morphs at three locations in the Caribbean and found that the nifH sequences recovered from M. cavernosa were consistent with previous studies on corals where members of both the α-proteobacteria and cyanobacteria were recovered. A number of nifH operational taxonomic units (OTUs) were significantly more abundant in the orange compared to the brown morphs, and one specific OTU (OTU 17), a cyanobacterial nifH sequence similar to others from corals and sponges and related to the cyanobacterial genus Cyanothece, was found in all orange morphs of M. cavernosa at all locations. The nifH diversity reported here, from a community perspective, was not significantly different between orange and brown morphs of M. cavernosa. Keywords  Montastraea cavernosa · Nitrogen fixation · nifH · Coral holobiont

Communicated by Joerg Overmann. Electronic supplementary material The online version of this article (doi:10.1007/s00203-013-0937-z) contains supplementary material, which is available to authorized users. N. D. Olson · M. P. Lesser (*)  Department of Molecular, Cellular and Biomedical Sciences, University of New Hampshire, Durham, NH 03824, USA e-mail: [email protected] N. D. Olson  Biochemical Sciences Division, National Institute of Standards and Technology, Gaithersburg, MD 20899, USA

Introduction Many environments, including oligotrophic coral reef ecosystems, depend on nitrogen fixation as a source of new nitrogen (Sohm et al. 2011; Zehr 2011), and many diazotrophic bacteria on coral reefs are found symbiotically in numerous taxa (Fiore et al. 2010; Kneip et al. 2007; O’Neil and Capone 2008). A large number of bacterial symbionts are known to be associated with scleractinian corals (Fiore et al. 2010; Rohwer et al. 2002), and many of them have been identified using metagenomic approaches as potentially being involved in nitrogen fixation and other transformations of nitrogen (Fiore et al. 2010; Kimes et al. 2010; Wegley et al. 2007). Previous studies have identified diazotrophic bacteria associated with the tissue of the coral species Acropora acuminate and Goniastrea australensis (Crossland and Barnes 1976) based on acetylene reduction assays, but did not investigate the identity of the organisms responsible for the observed nitrogen fixation. A subsequent study also used the acetylene reduction assay in combination with DCMU (3,4-dichlorophenyl dimethylurea), an inhibitor of photosystem II, to identify diazotrophic bacteria associated with the reef coral, A. variabilis (Williams et al. 1987). The inhibition of nitrogen fixation when exposed to DCMU indicated that the bacteria responsible for fixation were phototrophic, and most likely cyanobacteria. Finally, another study found photosynthesis-dependent nitrogen fixation associated with a number of different coral species and was able to culture a diazotrophic bacterium from a coral imprint using nitrogen-free media (Shashar et al. 1994). The identity of the cultured symbiont was then determined using southern hybridizations to be a γ-proteobacterial species, Klebsiella pneumoniae, indicating the presence of a heterotrophic diazotrophic symbiont.

13

854

Many studies now utilize culture-independent PCRbased techniques to quantify the diverse community of bacteria associated with reef-building corals, including species known to fix nitrogen. Most of these diazotrophic bacteria belonged to the Phylum Cyanobacteria (Fiore et al. 2010), and a recent metagenomic analysis of the coral Porites astreoides identified genes involved in nitrogen fixation, including nifH, belonging to the Phylum Cyanobacteria (Wegley et al. 2007). Additionally, a microarray (geoChip 2.0) analysis of Montastraea faveolata revealed a diverse community of both heterotrophic and autotrophic diazotrophic bacteria (Kimes et al. 2010), while nifH clone libraries from the Hawaiian corals Montipora capitata and M. flabellata also identified a diverse group of heterotrophic and autotrophic diazotrophic bacteria (Olson et al. 2009). The most recent work available using nifH sequences suggests that there may be characteristic associations of diazotrophic bacteria with several species of scleractinian coral from the Great Barrier Reef and that many of these are closely related to the bacterial group rhizobia (Lema et al. 2012). One system that has received attention regarding the presence and function of diazotrophic symbionts is the Caribbean coral M. cavernosa (Lesser et al. 2004, 2007). Using multiple techniques, cyanobacterial endosymbionts were found to be associated with the tissues of orange-colored colonies of this species, but not brown/green colonies (Lesser et al. 2004). One symbiont was found in abundance (107 cells cm−2) and was a member of the genus Synechococcus based on 16S rRNA sequence analysis in orange but not brown morphs of this coral (Lesser et al. 2004). Additionally, acetylene reduction measurements showed that nitrogen fixation occurs only in the orange color morph in a diel pattern with peak fixation at dawn and dusk (Lesser et al. 2007). These findings suggested that the presence of this cyanobacterium was responsible for the differences in nitrogen fixation between the two color morphs. However, the diversity of nitrogen fixing bacteria associated with the two color morphs has never been investigated. Here, we assess and compare the diversity of diazotrophic bacteria between the two color morphs of M. cavernosa at three locations within the Caribbean basin using nifH clone libraries (Zehr et al. 2003) that codes for a subunit of the nitrogenase enzyme which is responsible for nitrogen fixation.

Methods Coral sample collection and processing Corals samples were collected from a depth of ~15 m using SCUBA where populations of the two color morphs co-occur in high numbers at all locations. Both orange

13

Arch Microbiol (2013) 195:853–859

(N  = 3) and brown (N  = 3) M. cavernosa colonies were sampled from Alligator Reef, Florida Keys, USA (24°51′N, 80°36′W), Rock Bottom Wall, Little Cayman Island (19°18′N, 81°16′W), and North Perry Reef, Lee Stocking Island, Bahamas (23°46′ N, 76°05′ W), for a total of 18 colonies. All samples from all locations were collected at ~noon and had similar temperatures of 25–28 °C and downwelling irradiances (maximum photosynthetically active radiation; PAR: 400–700 nm) of ~500–600 μmol quanta m−2 s−1 (Lesser unpublished). Samples were transported in coolers with seawater to shore where corals were initially processed. First, loosely associated bacteria are removed from corals by holding them upside down and using low-pressure airbrushing with sterile (0.22 μm) filtered seawater to induce the production and loss of copious amounts of mucous and mucous-associated microbes (Lesser et al. 2004; Olson et al. 2009). Individual polyps were then removed from the sample using bone cutters, preserved in DNA buffer and frozen at −50 °C (Seutin et al. 1991). Samples were transported to the University of New Hampshire on ice for processing, and samples did not thaw during transport. Finally, for Lee Stocking Island, orange and brown colonies of M. cavernosa collected from the same depth represent genetically distinct populations as assessed by amplified fragment length polymorphisms (AFLP) and that the communities of symbiotic Symbiodinium sp. phylotypes are not significantly different between color morphs (Jarrett and Lesser, unpublished). DNA extraction, PCR amplification, and cloning Genomic DNA was extracted from the samples using a cetyltrimethylammonium bromide (CTAB) protocol (France and Kocher 1996), and DNA concentrations were determined using a NanoDrop spectrophotometer (Thermo Scientific, Waltham MA.) at 260 nm. The following protocol was used to amplify the nifH gene from the extracted genomic DNA. The most common protocol used to recover nifH sequences is a nested protocol described by Zehr and McReynolds (1989) that results in a 359 bp product. Several other universal primer sets have also been developed with varying degrees of degeneracy that can result in mismatches and taxonomic biases for the sequences recovered from any particular sample (Gaby and Buckley 2012). Using an in silico analysis, 19 universal primer sets were compared using nifH sequences from publically available databases which included sequences from all four identified nifH taxonomic clusters (Olson 2010). As a result of that analysis, the primer set YAA (5′-ATR TTR TTNGCN GCR TA-3′) and IGK (5′-AAR GGN GGN ATH GGN AA-3′) (Ohkuma and Kudo 1996) was utilized for the initial PCR reaction (462 bp product) because it targets the same area as the nifH4 primer (Zehr and McReynolds

Arch Microbiol (2013) 195:853–859

1989) with fewer mismatches and greater universality, and nifH1 (5′-TGY GAY CCN AAR GCN GA-3′) and nifH2 (5′-AND GCC ATC ATY TCN CC-3′) from the Zehr and McReynolds (1989) protocol for the second PCR reaction of the nested PCR resulting in products of ~359 bp. The first round reaction consisted 0.5 × TITANIUM Taq DNA Polymerase and 1 X Buffer (Clontech, Mountain View, CA USA), 0.2 μM dNTP’s, 0.4 μM forward and reverse primer, and 1 μl of template DNA with a total volume of 25 μl. The cycling protocol consisted of a 3 min initial denaturation step at 95 °C followed by 10 cycles of a annealing step at 53.5 and a 72 °C extension step all of which were 30 s followed by a 7-min final extension step at 72 °C. The second round reaction consisted of 0.5 × TITANIUM Taq DNA Polymerase and 1 × Buffer (Clontech, Mountain View, CA USA), 0.2 μM dNTP’s, 0.8 μM forward and reverse primer, and 2 μl of the initial PCR product in a total volume of 25 μl. The same cycle protocol as the first round was employed except that a 10 cycle reaction was performed instead of a 30 cycle reaction as more commonly employed (Zehr and McReynolds 1989). Second round PCR products were gel-extracted and cloned as previously described for diazotrophs from corals (Olson et al. 2009). For all PCR runs, negative controls (i.e., no template on first or second reactions) were employed and no amplification was observed. Sequence analysis Sequences were trimmed and filtered using the Ribosomal Database Pipeline (http://rdp.cme.msu.edu/), compared to the GenBank database using blastn as well as an extensive, and publically available, nifH database (pmc.ucsc.edu/~wwwzehr/research/database/) using the ARB software package (v 5.2) to confirm that all sequences were from the nifH gene. The open reading frames of the nifH sequences (~249 bp) from all clones were translated into their amino acid sequences using the bacterial codon table and aligned in MEGA4 using the clustalW algorithm and a distance matrix calculated using the PHYLIP software package (http://evolution.genetics.washington.edu/phylip.html). The distance matrix was used to assign operational taxonomic units (OTUs at 95 % sequence homology) from which richness estimates (Chao1) and diversity indices (Shannon-Weiner) were calculated using Mothur v 1.11.0 (www.mothur.org). Sample coverage was determined by dividing the number of obtained OTUs by the Chao1 estimate for the sample. The OTU abundances for pooled samples were calculated and statistically compared using metastats (http://metasta ts.cbcb.umd.edu/). The aligned nifH sequences were used to construct a neighbor joining phylogenetic tree with 1,000 bootstrap replicates, and then, a whole community

855

analysis with UniFrac (http://bmf2.colorado.edu/unifrac/ index.psp) was conducted. All nifH sequences obtained in this study were submitted to GenBank (accession number HM999081-HM999629).

Results and discussion From clone libraries of the nifH gene, we were able to compare the communities of diazotrophic bacteria associated with the Caribbean coral M. cavernosa. We quantified the diversity of nifH sequences between color morphs at geographically separated locations and identified a number of specific nifH OTUs in the population of orange morphs that would be consistent with previous studies using 16S rRNA as a marker (Lesser et al. 2004) and might account for the observed differences in nitrogen fixation between the two color morphs (Lesser et al. 2007). A total of 515 nifH sequences were obtained from the 18 coral samples. The number of clones sequenced per coral sample ranged from 10 to 33, and rarefaction curves for both brown and orange morphs at all locations are curvilinear in nature showing that sampling effort was almost saturated using the nested primer set described above (Fig. 1). A diverse community of diazotrophic bacteria dominated by proteobacteria was found in association with M. cavernosa (STable 1). For each of the coral samples, the number of observed OTUs, Chao1 richness estimate, Shannon-Weiner diversity index, and percent coverage was determined (STable 2). The Chao1 richness estimates varied between corals with values ranging from 2 to 32 OTUs based on a 95 % sequence similarity cutoff. These estimates were often greater than the observed OTUs indicating that despite the curvilinear nature

Fig. 1  Rarefaction curves for sequences pooled by location and color at a 0.05 similarity cutoff. Solid line is Florida Keys, short dash is Little Cayman Island, and long dash is Lee Stocking Island

13

856

of the rarefaction curves the libraries probably underestimate total nifH diversity in these samples. The depth of the sequencing effort is reflected in the percent coverage with values ranging from 41.2 to 100 %. In addition to percent coverage, the Shannon-Wiener index of diversity (H) was calculated for nifH sequences between color morphs and geographic locations with higher H values indicating greater nifH sequence diversity. Diversity was always greater for samples having higher observed and estimated numbers of OTUs (STable 2) as expected. The taxonomic breadth of nifH sequences recovered from M. cavernosa is also consistent with previous studies that investigated the diversity of diazotrophic bacteria associated with reef-building corals as is the dominance of OTUs represented by members of the α-proteobacteria and cyanobacteria (Olson et al. 2009; Kimes et al. 2010; Lema et al. 2012). Fig. 2  Color and location comparison for the percent abundance for individual OTUs. Comparison of orange and brown samples pooled for all locations (a), Florida Keys (b), Little Cayman Island (c), Lee Stocking Island (d). Asterisks indicate OTUs where the percent abundance is statistically different between color morphs for pooled samples

13

Arch Microbiol (2013) 195:853–859

For the comparison of nifH sequences recovered from the orange versus brown morphs, our OTU groupings were set at a sequence similarity cutoff of 95 %, a cutoff commonly used in other nifH diversity studies and representing a species to genus taxonomic grouping (Hsu and Buckley 2009; Konstantinidis and Tiedje 2005). A total of 52 OTUs were identified for both color morphs with the closest relative of each OTU determined using the tBlastn algorithm (STable 1). Based on the statistical analysis of the mean percent abundance, 7 OTUs were found statistically different between the two color morphs when all samples were pooled (Fig. 2a). Of these seven, five had a greater mean abundance in orange versus brown color morphs. These include OTUs 35, 37, and 52 that belong to the bacterial class γ-proteobacteria and OTU 17 and 45 that are in the Phylum Cyanobacteria and class δ-proteobacteria,

Arch Microbiol (2013) 195:853–859

857

respectively (STable 1). The two OTUs that were more abundant for the brown color morph were OTU 1 and OTU 6 that were a cyanobacterium and a α-proteobacterium, respectively. Both were identified as nifH sequences, with 100 and 97 % identity, respectively, to sequences recovered from the coral M. flabellata from Hawaii (Olson et al. 2009). In the Florida Keys, no OTUs were significantly

different between orange and brown morphs (Fig. 2b), while in Little Cayman Island, OTUs 2, 37, and 50 were significantly greater in orange morphs (Fig. 2c). OTU 2 belongs to the class α-proteobacteria, while OTUs 37 and 50 belong to the class γ-proteobacteria (STable 1). For Lee Stocking Island, OTU 19, a α-proteobacterium, was more abundant in brown colonies and OTU 35, a γ-proteobacterium, was

Fig. 3  Whole community analysis for individual samples and pooled samples. Sample identification indicated by the first letter indicating color orange (O) or brown (B), followed by location Florida Keys (F), Little Cayman Island (LC), or Lee Stocking Island (LS), with replicate samples indicated by a number 1–3. Graphs depict the first three

principal component analyses of individual replicate samples based on UniFrac distance calculations (a–c). The clustering of pooled communities by location and color is based on UniFrac distance calculations. The average linkage (UMPGA) clustering based on UniFrac distance estimates for pooled replicates by location and color (d)

13

858

more abundant in orange colonies (Fig. 2d). OTU 17, a cyanobacterium, was present in all orange morphs from all locations and in significantly higher abundance for orange morphs when all three locations are pooled. Additionally, it was never found in brown morphs of M. cavernosa from Little Cayman Island or Lee Stocking Island (Fig. 2c, d). OTU 17 nifH sequences from M. cavernosa were most similar to sequences obtained from coral reef sponges in the cyanobacterial genus Cyanothece (Mohamed et al. 2008) that is closely related to cyanobacteria in the genus Synechococcus. These genera are highly diverse and polyphyletic containing both nitrogen fixing and non-nitrogen fixing representatives (Bandyopadhyay et al. 2011; Robertson et al. 2001). The original sequence identified as a cyanobacterial symbiont in M. cavernosa using 16S rRNA primers (Lesser et al. 2004) and identified as a member of the genus Synechococcus at that time is also 98 % similar to multiple sequences identified as Synechococcus spongiarum originally described as a symbiont of sponges (Robertson et al. 2001). At least in the case of this OTU, it suggests some congruence between the nifH OTU 17 and the 16S rRNA sequence from the cyanobacterial symbiont originally described from the orange morphs of M. cavernosa. We were also interested in assessing differences in diazotrophic bacterial communities between individual corals and between locations. Using UniFrac and a principal coordinate analysis, the nifH communities did not group by either color or location (Fig. 3a–c). However, when pooled by location and color, they cluster together by location rather than color (Fig 3d) with Florida Keys and Little Cayman samples clustering more closely to each other and being distinct from Lee Stocking Island samples. However, differences in the pooled groups reflected in the clustering were not significant as determined using the UniFrac significance test. These results suggest that the taxonomic composition of the communities do not, a priori, explain the functional differences in nitrogen fixation between the orange and brown color morphs. One of the most interesting findings from this study is that the OTU representing the greatest number of recovered sequences in this study (i.e., OTU 19) was identified as a member of the rhizobia (Bradyrhizobium jicamae, STable 1) as in the study of Acropora sp. and Pocilliopora damicornis from the Great Barrier Reef (Lema et al. 2012) and that many of the common nifH sequences are unidentified. While other studies of diazotroph diversity in corals have recovered rhizobia sequences (Olson et al. 2009), the dominant heterotrophic nitrogen fixer in corals is more commonly a species of Vibrio (Chimetto et al. 2008; Olson et al. 2009). Taken together, the diazotrophic communities described here from the orange and brown morphs of M. cavernosa could not be statistically distinguished from each other, but it does not eliminate the possibility that the difference

13

Arch Microbiol (2013) 195:853–859

in rates of nitrogen fixation between the two color morphs may be due to multiple OTUs in significantly greater percent abundance in orange morphs, or to a specific OTU (e.g., OTU 17). Not assessed in this study were the absolute numerical differences in the total number of nitrogen fixing bacteria represented by the OTUs found in the orange versus the brown morph. Additionally, because of the multicopy nature of nifH (Church et al. 2005) and the fact that we do not know which members of this community are actively fixing nitrogen, nifH should also be analyzed at the expression level. As an example of how informative this could be, a study by Mohamed et al. (2008) investigated the diversity of nitrogen fixing bacteria associated with coral reef sponges. In that study, they found a similar diversity of nitrogen fixing bacteria associated with sponges as we have described here for corals, but at the expression level, transcripts were exclusively cyanobacterial in origin. In order to understand and quantify the importance of nitrogen fixation in the nutrient biogeochemistry of M. cavernosa, or any scleractinian coral, a comprehensive molecular approach will be required to simultaneously assess the taxonomy, numerical abundance, tissue-specific location, and expression patterns of nitrogen fixing bacteria. We are currently analyzing comparative meta-transcriptomes of both the orange and brown morphs of M. cavernosa using Illumina paired-end sequencing that should make significant progress in linking taxonomy and function. Acknowledgments This research was supported by grants from the National Science Foundation, National Geographic Society, the National Oceanic and Atmospheric Administration National Institute of Undersea Science and Technology, the University of New Hampshire Marine Program, and the American Museum of Natural History Lerner-Grey Fellowship for Marine Research. The work presented here was submitted as a part of the requirements for the completion of Nathan Olson’s Master’s degree to the Graduate School at the University of New Hampshire. This manuscript was improved by constructive comments from Anton Post.

References Bandyopadhyay A, Elvitigala T, Welsh E, Stöckel J, Liberton M, Min H, Sherman LA, Pakrasi HB (2011) Novel metabolic attributes of the genus Cyanothece, comprising a group of unicellular nitrogen-fixing cyanobacteria. mBio 2:e00214-11 Chimetto LA, Brocchi M, Thompson CC, Martins RCR, Ramos HR, Thompson FL (2008) Vibrios dominate as culturable nitrogenfixing bacteria of the Brazilian coral Mussismilia hispida. Syst Appl Microbiol 31:312–319 Church MJ, Short CM, Jenkind BD, Karl DM, Zehr JP (2005) Temporal patterns of nitrogenase (nifH) expression in the oligotrophic North Pacific Ocean. Appl Environ Microbiol 71:5362–5370 Crossland CJ, Barnes DJ (1976) Acetylene reduction by coral skeletons. Limnol Oceanogr 21:153–156 Fiore CL, Jarett JK, Olson ND, Lesser MP (2010) Nitrogen fixation and nitrogen transformations in marine symbioses. Trends Microbiol 18:455–463

Arch Microbiol (2013) 195:853–859 France SC, Kocher TD (1996) DNA sequencing of formalin fixed crustaceans from archival research collections. Mol Mar Biol Biotech 5:304–313 Gaby JC, Buckley DH (2012) A comprehensive evaluation of PCR primers to amplify the nifH gene of nitrogenase. PLoS ONE 7:e42149 Hsu S-F, Buckley DH (2009) Evidence for the functional significance of diazotroph community structure in soil. ISME J 3:124–136 Kimes NE, Van Nostrand JD, Weil E, Zhou J, Morris PJ (2010) Microbial functional structure of Montastraea faveolata, an important Caribbean reef-building coral, differs between healthy and yellow-band diseased colonies. Environ Microbiol 12:541–556 Kneip C, Lockhart P, Voß C, Maier U-G (2007) Nitrogen fixation in eukaryotes—New models for symbiosis. BMC Evol Biol 7:55 Konstantinidis KT, Tiedje JM (2005) Genomic insights that advance the species definition for prokaryotes. Proc Natl Acad Sci 102:2567–2572 Lema KA, Willis BL, Bourne DG (2012) Corals form characteristic associations with symbiotic nitrogen-fixing bacteria. Appl Environ Microbiol 78:3136–3144 Lesser MP, Mazel CH, Gorbunov MY, Falkowski PG (2004) Discovery of symbiotic nitrogen-fixing cyanobacteria in corals. Science 305:997–1000 Lesser MP, Falcón LI, Rodríguez-Román A, Enríquez S, Hoegh-Guldberg O, Iglesias-Prieto R (2007) Nitrogen fixation by symbiotic cyanobacteria provides a source of nitrogen for the scleractinian coral Montastraea cavernosa. Mar Ecol Prog Ser 346:143–152 Mohamed NM, Colman AS, Tal Y, Hill RT (2008) Diversity and expression of nitrogen fixation genes in bacterial symbionts of marine sponges. Environ Microbiol 10:2910–2921 Ohkuma M, Kudo T (1996) Phylogenetic diversity of the intestinal bacterial community in the termite Reticulitermes speratus. Appl Environ Microbiol 62:461–468 Olson ND (2010) nifH diversity associated with Montastraea cavernosa identified using an optimized primer protocol. M.S. thesis. University of New Hamshire, Durham, NH Olson ND, Ainsworth TD, Gates RD, Takabayashi M (2009) Diazotrophic bacteria associated with Hawaiian Montipora corals:

859 diversity and abundance in correlation with symbiotic dinoflagellates. J Exp Mar Biol Ecol 371:140–146 O’Neil JM, Capone DG (2008) Nitrogen cycling in coral reef environments. In: Bronk DA, Mullholland MR, Carpenter EJ (eds) Nitrogen in the marine environment. Elsevier, Capone, pp 949–989 Robertson BR, Tezuka N, Watanabe MM (2001) Phylogenetic analyses of Synechococcus strains (cyanobacteria) using sequences of 16S rDNA and part of the phycocyanin operon reveal multiple evolutionary lines and reflect phycobilin content. Int J Syst Evol Microbiol 51:861–871 Rohwer F, Seguritan V, Azam F, Knowlton N (2002) Diversity and distribution of coral-associated bacteria. Mar Ecol Prog Ser 243:1–10 Seutin G, White BN, Boag PT (1991) Preservation of avian blood and tissue samples for DNA analysis. Can J Zool 69:82–90 Shashar N, Cohen Y, Loya Y, Sar N (1994) Nitrogen fixation (acetylene reduction) in stony corals: evidence for coral-bacteria interactions. Mar Ecol Prog Ser 111:259–264 Sohm JA, Webb EA, Capone DG (2011) Emerging patterns of marine nitrogen fixation. Nat Rev Microbiol 9:499–508 Wegley L, Edwards R, Rodriguez-Brito B, Liu H, Rohwer F (2007) Metagenomic analysis of the microbial community associated with the coral Porites astreoides. Environ Microbiol 9:2707–2719 Williams WM, Viner AB, Broughton WJ (1987) Nitrogen fixation (acetylene reduction) associated with the living coral Acropora variabilis. Mar Biol 94:531–535 Zehr JP (2011) Nitrogen fixation by marine cyanobacteria. Trends Microbiol 19:162–173 Zehr JP, McReynolds LA (1989) Use of degenerate oligonucleotides for amplification of the nifH gene from the marine cyanobacterium Trichodesmium thiebautii. Appl Environ Microbiol 55:2522–2526 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

13