ORIGINAL ARTICLE

Successional changes in bacterialcommunities during the development ofblack band disease on the reef coral,Montipora hispida

Yui Sato1,2, Bette L Willis1 and David G Bourne2

1ARC Centre of Excellence for Coral Reef Studies and School of Marine and Tropical Biology, James CookUniversity, Townsville, Queensland, Australia and 2Australian Institute of Marine Science (AIMS),Townsville MC, Townsville, Queensland, Australia

Black band disease (BBD) consists of a mat-forming microbial consortium that migrates acrosscoral colonies causing rapid tissue loss. Although BBD-associated microbial communities havebeen well characterized, little is known regarding how these complex bacterial consortia develop.This study analyzed successional changes in microbial communities leading to the development ofBBD. Long-term monitoring of tagged corals throughout outbreaks of BBD in the central GreatBarrier Reef documented cyanobacterium-infected lesions, herein termed cyanobacterial patch(es)(CP), which were macroscopically distinct from BBD and preceded the onset of BBD in 19% of thecases. Dominant cyanobacteria within CP lesions were morphologically distinct from onesdominating BBD lesions. Clone libraries and terminal restriction fragment length polymorphismanalysis confirmed shifts within cyanobacterial assemblages, from Blennothrix sp.-affiliatedsequences dominating CP lesions, to Oscillatoria sp.-affiliated sequences, similar to those retrievedfrom other BBD samples worldwide, dominating BBD lesions. Bacterial 16S ribosomal RNA clonelibraries also showed shifts in bacterial ribotypes during transitions from CP to BBD, withAlphaproteobacteria-affiliated sequences dominant in CP libraries, whereas gammaproteobacterialand cyanobacterial ribotypes were more abundant in BBD clone libraries. Sequences affiliated withorganisms identified in sulfur cycling were commonly retrieved from lesions showing characteristicfield signs of BBD. As high sulfide concentrations have been implicated in BBD-mediated coraltissue degradation, proliferation of a microbial community actively involved in sulfur cyclingpotentially contributes to the higher progression rates found for BBD compared with CP lesions.Results show how microbial colonization of indistinct lesions may facilitate a common coral diseasewith proven ecological effects on coral populations.The ISME Journal (2010) 4, 203–214; doi:10.1038/ismej.2009.103; published online 24 September 2009Subject Category: microbe-microbe and microbe-host interactionsKeywords: black band disease (BBD); coral; bacteria; cyanobacteria; transition

Introduction

Black band disease (BBD), which manifests as adarkly pigmented microbial mat that migrates acrosslive coral colonies causing necrosis of coral tissues(reviewed in Richardson, 2004), is a major contri-butor to ongoing loss of reef building corals world-wide, particularly in the Caribbean (Bruckner and

Bruckner, 1997; Goreau et al., 1998; Green andBruckner, 2000; Sutherland et al., 2004; Kaczmarsky,2006; Page and Willis, 2006; Sato et al., 2009). BBDhas been termed a poly-microbial disease because aprimary causative agent(s) has not been identifiedand a tightly organized, complex microbial consor-tium appears to act in concert to cause coral tissuenecrosis (Richardson et al., 1997, 2007; Richardson,2004). Currently, however, little is known regardinghow bacterial pathogens form the black band on hostcorals or how changes in microbial communitieslead to onset of the disease. Given the potential forBBD to cause substantial tissue loss in coralpopulations (reviewed in Green and Bruckner,2000; Sato et al., 2009) and the likelihood thateffects of the disease will be exacerbated as seawater

Received 6 July 2009; revised 24 August 2009; accepted 24 August2009; published online 24 September 2009

Correspondence: Y Sato, ARC Centre of Excellence for Coral ReefStudies and School of Marine and Tropical Biology, James CookUniversity, 101 Angus Smith Drive, Townsville, Queensland4811, Australia.E-mail: yui.sato1@jcu.edu.au

The ISME Journal (2010) 4, 203–214& 2010 International Society for Microbial Ecology All rights reserved 1751-7362/10 $32.00

www.nature.com/ismej

temperatures warm with predicted climate change(Bruckner et al., 1997; Kuta and Richardson, 2002;Borger and Steiner, 2005; Voss and Richardson,2006; Rodriguez and Croquer, 2008; Sato et al.,2009), there is need for greater understanding ofearly stages in the onset of the disease if BBD is to bemanaged effectively.

It is well-established that the microbial consor-tium making up the black band is dominated bycyanobacteria, but also includes sulfate-reducingbacteria, sulfide-oxidizing bacteria, marine fungiand other heterotrophic microbes (reviewed inRichardson, 2004). Molecular analyses of BBD-dominating cyanobacteria have retrieved bacterial16S ribosomal RNA (rRNA) gene sequences af-filiated with many cyanobacterial species/strains,such as a Geitlerinema sp. (Cooney et al., 2002;Myers et al., 2007) (formerly known as Phormidiumcorallyticum), Trichodesmium spp. (Frias-Lopezet al., 2002, 2003), Leptolyngbya sp. (Myers et al.,2007) and an Oscillatoria sp., which has been mostcommonly detected in BBD samples collected fromthe wider Caribbean, Indo-Pacific and the Red Sea(reviewed in Myers et al., 2007; Myers and Richard-son, 2009). To what extent each of these cyanobac-terial strains is involved in BBD pathogenesis isunclear, however, there is emerging evidence thatBBD-associated cyanobacteria vary between geogra-phical and regional locations and host coral species(Voss et al., 2007), yet each may occupy the sameecological niche in BBD microbial mats (Myers andRichardson, 2009).

Pathogenesis of BBD is currently thought toinvolve light-associated, vertical microgradients inoxygen and sulfide (Carlton and Richardson, 1995),accompanied by down-migration of the filamentouscyanobacteria when light levels are high (Richard-son, 1996; Viehman and Richardson, 2002). Subse-quent cyanobacterial penetration of coral tissue(Barneah et al., 2007) suggests that mechanicaland/or chemical degeneration of coral tissue is alsoinvolved in pathogenicity. In addition, genes for thebiosynthesis of microcystin have been detected fromGeitlerinema sp. and Leptolyngbya sp. derived fromBBD samples, and production of the toxin has beenconfirmed with high-performance liquid chromato-graphy/mass spectrometry and a protein phospha-tase inhibition assay (Richardson et al., 2007). Thesefindings indicate that the cyanotoxin may contributeto coral tissue necrosis caused by BBD. Desulfovi-brio spp. are also thought to have a role in BBDpathogenesis through production and accumulationof sulfide, which is present at high concentrationsunder the BBD mat and can be lethal to coral tissuewhen combined with the anoxic microenviron-ment (Richardson et al., 1997). Sequences affiliatedwith other potential pathogenic bacteria, includingCytophaga sp., Clostridium sp., Campylobacter sp.,Arcobacter sp. and an alphaproteobacterium respon-sible for juvenile oyster disease, have been retrievedfrom Caribbean BBD samples (Cooney et al., 2002;

Frias-Lopez et al., 2002), although it is unknownwhether these organisms have any role in thepathogenesis of BBD.

Little is known regarding how bacterial pathogensform the characteristic black band on host coralsor regarding microbial communities during theearly onset of BBD in nature. In one publishedobservation, a new BBD infection on Diploriastrigosa was documented as a small pigmented ringa few centimeter in diameter, which apparently hada BBD microbial mat already formed (Antonius,1981). Given that BBD-associated bacteria may bepresent on or within healthy coral tissues at back-ground levels (Frias-Lopez et al., 2002), the diseaseband may be formed when environmental factorschange and/or host coral health is biologically orphysically compromised (Rohwer et al., 2002).For example, the BBD-associated cyanobacteriumdesignated as CD1C11 is commonly detected inhealthy tissues of Montastraea annularis withculture-independent methods (Klaus et al., 2007).It has also been shown that predation of octo-corals by corallivorous animals (for example, thegastropod Cyphoma gibbosum and polychaeteHermodice carunculata) facilitates BBD infectionwhere BBD is present (Antonius, 1985). Lack ofknowledge regarding transmission modes andearly developmental stages of the BBD microbialconsortium hinders development of reef manage-ment strategies to mitigate the impacts of thisdisease.

Recently, recurring summer outbreaks of BBDwere documented in a long-term monitoring studyof tagged corals in an assemblage of Montiporaspecies on an inshore reef in the central GreatBarrier Reef (GBR) (Sato et al., 2009). In some cases,actively expanding cyanobacterium-infected lesions,whose green or brown colorations were distinctfrom characteristic BBD lesions, were identifiedbefore the onset of BBD (Figure 1). We herein referto these lesions as ‘cyanobacterial patch(es)’ or‘CP’. Preliminary microscopic observations of CPindicate that they contain cyanobacteria that aremorphologically distinct from ones observed insubsequent BBD lesions, suggesting that CP andBBD have different overall microbial communities.In this study, we analyze successional changes inmicrobial communities from early CP lesions tofully developed BBD lesions to document develop-ment of the complex bacterial consortia thatcharacterize BBD. Seasonal patterns in abundanceof CP and BBD were documented to determinehow often and when CP microbial communitiesdevelop into BBD microbial consortia. We comparethe virulence of CP and BBD by recording progres-sion rates of the lesions on host corals and byhistological observations. In addition, we determinephylogenetic diversity of bacteria associated withCP and BBD lesions to examine changes in thesemicrobial communities during the development ofBBD from CP.

Bacterial transition in the development of BBDY Sato et al

204

The ISME Journal

Materials and methods

Study site and field surveysMontipora assemblages on an inshore reef at PelorusIsland in the central GRB were monitored betweenSeptember 2006 and January 2009 to examineseasonal patterns in prevalences of CP and BBD.The study site is exposed to minimal levels ofterrestrial run-off or human impact (see Sato et al.,2009 for a detailed description of the study site). Inaddition to the three previously described quadratsat the southeast corner of the island (18133.6’S,146130.1’E), three replicate 10 m� 10 m permanentquadrats were established at the northeast cornerof the island (18132.5’S, 146130.0’E) in the samemanner in May 2007. CP-infected colonies wereindividually tagged and monitored approximatelymonthly to determine whether CP lesions devel-oped into BBD. Progression rates of CP and BBDlesions were measured throughout the study periodas described previously (Sato et al., 2009), andannual means and means for summer (December2007BFebruary 2008) and winter (June 2007BAugust 2007) were statistically compared betweendisease signs and seasons.

Sample collectionSamples of CP and BBD lesions were collectedfrom five tagged colonies of Montipora hispidaadjacent to the monitoring plots at depths of 2.5–3.0 m between August 2007 and February 2008. Forbacterial examination, samples were collected atapproximately 2-week intervals to monitor changesin microbial communities as CP lesions progressedinto characteristic BBD lesions. To minimize sam-pling effects on disease progression, a small (ap-proximately 5 mm in diameter and 2 mm in depth)portion of the cyanobacterial mat and skeletonunderneath the mat was collected from the front ofeach lesion, using a separate sterilized stainless steelchisel for each sample to avoid cross contamination.The samples were placed in individual 1.5 ml screwcap tubes underwater and kept on ice upon return-ing to the surface. After examining the cyanobacter-ial morphology under a phase contrast microscope(BX41TF, Olympus Corporation, Tokyo, Japan) at amagnification of 400 times, excess seawater wasremoved and lesion samples were preserved in100% ethanol at �20 1C. A small portion of coraltissue with attached skeleton was also sampled fromtwo healthy colonies and preserved in the samemanner. For histological observations, small(3 cm� 3 cm) diseased coral fragments weresampled from each of five CP- and BBD-infectedcolonies and immediately fixed in 10% seawaterformalin. In addition, an ambient seawater sample(1 l) was collected from approximately 30 cm abovethe sampled colonies and immediately filteredthrough a 0.22 mm Millipore Sterivex filter unit(Sigma-Aldrich, St Louis, MO, USA). The filter unit

Figure 1 Transition in macroscopic signs of a disease lesion atthe identical site on a colony of Montipora hispida (colony 3) from:(a) a typical CP (7 October 2007), through (b) an intermediate stagebetween CP and BBD (25 October 2007), to (c) a well-developedBBD lesion (10 November 2007). Abbreviations: BBD, black banddisease; CP, cyanobacterial patch; IM, intermediate stage betweenCP and BBD.

Bacterial transition in the development of BBDY Sato et al

205

The ISME Journal

was kept on ice while transported to the laboratorybefore being stored at �20 1C.

Histological preparationFormalin fixed samples were placed betweensponge pads in histology cassettes to avoid detach-ment of the disease mat from the underlying coraltissue during decalcification in a 1 to 3% series offormic acid washes. After paraffin wax embeddingof decalcified samples, transverse sections of coraltissue (5 mm thickness) from the edge of each lesionwere prepared. Sections were stained with hema-toxylin and eosin to examine the extent of tissuestructure degradation.

DNA extractionDNA extraction from ethanol preserved lesions andcoral tissue samples was performed with the Power-Plant DNA Isolation Kit (MO BIO Laboratories,Carlsbad, CA, USA) as described by the manufac-turer with the following modifications. After theinitial preparation, 2� 30-s bead-beating cycles witha 30-s interval were performed with a Mini-Bead-beater (Biospec Products, Bartleville, OK, USA).DNA from the seawater filter unit was isolatedaccording to the methods outlined in Schauer et al.(2000). The quality of extracted DNA was verified on1% agarose gel stained with ethidium bromide andquantified with a NanoDrop Spectrophotometer(Thermo Fisher Scientific, Waltham, MA, USA).

PCR amplification, cloning and sequencingCyanobacterial clone libraries were constructedfrom microbial mat samples derived from two taggedcorals, each with characteristic CP lesions (colonies1 and 3, sampled on 24 August and 7 October,respectively). Lesions on these two colonies wererepeatedly sampled as they visually transitionedinto characteristic BBD lesions (fully developedBBD samples were collected on 22 September and10 November, respectively). Cyanobacteria-specificPCR primers CYA106F (CGGACGGGTGAGTAACGCGTGA) and CYA781R (an equimolar mixture of CYA781R(a) (GACTACTGGGGTATCTAATCCCATT) andCYA781R(b) (GACTACAGGGGTATCTAATCCCTTT))(Nubel et al., 1997) were used to amplify an approxi-mately 700-bp region of the 16S rRNA gene. Eachreaction mixture of 50 ml contained 0.2 mM of eachprimer, 0.2 mM of each dNTP, 1�PCR ReactionBuffer (TAQ DNA Polymerase Kit, Scientifix, Clayton,Australia), 0.08% (w/v) bovine serum albumin,1.25 U of TAQ DNA Polymerase (Scientifix) and10–20 ng of template DNA. Amplification was per-formed with initial melting at 95 1C for 5 min,followed by 35 cycles of 94 1C for 1 min, 60 1C for1 min and 72 1C for 1 min, and a final extension at72 1C for 10 min. PCR products were purified withgel extraction using the QIAquick Gel Extraction Kit

(Qiagen, Hilden, Germany) and cloned with the T&ACloning Kit (Real Biotech Corporation, Taipei,Taiwan) and TOP10F’ Competent Cells (Invitrogen,Carlsbad, CA, USA). A total of 72 clones were ran-domly selected from each library and the insertedproduct was PCR re-amplified with the M13 primerssupplied with the T&A Cloning Kit. Operationaltaxonomic units (OTUs) were identified with re-striction fragment length polymorphism usingenzymes MspI and RsaI (New England BioLabs,Beverly, MA, USA). Clones representative of eachOTU group were sequenced with the M13 forwardprimer (Macrogen Inc., Seoul, Korea). A cyanobac-terial 16S rRNA gene clone library was alsoconstructed from seawater sampled above the coralcolonies using the method outlined above. Cloningand sequencing of a total of 96 randomly selectedclones of this sample were performed at AustralianGenome Research Facility Ltd (Brisbane, Australia).

Bacterial 16S rRNA gene clone libraries wereconstructed from microbial samples derived fromthree coral colonies (colonies 1, 2 and 3), each with acharacteristic CP lesion that was repeatedly sampledthroughout its transition into BBD (lesions weresampled from 24 August to 22 September, 7 Octoberto 10 December and 24 August to 23 November,respectively). PCR amplification of an approxima-tely 1300-bp region was performed with bacterial-specific forward primer 63F (CAGGCCTAACACATGCAAGTC) and reverse primer 1387R (GGGCGGWGTGTACAAGGC) (Marchesi et al., 1998). Eachreaction mixture of 50 ml contained 0.2 mM of eachprimer, 0.2 mM of each dNTP, 1�Buffer (HotStarTaqDNA Polymerase Kit, Qiagen), 1 U of HotStarTaqDNA Polymerase (Qiagen) and 2–10 ng of templateDNA. Amplification was performed with initialheating at 95 1C for 15 min, followed by 30 cyclesof 94 1C for 1 min, 55 1C for 1 min and 72 1C for 1 min,followed by a final extension step at 72 1C for10 min. PCR products were purified with theQIAquick Gel Extraction Kit (Qiagen) and ligatedinto the TOPO TA cloning vector (Invitrogen).Cloning and sequencing of a total of 96 randomclones per library using the M13 forward primerwere performed by the Australian Genome ResearchFacility Ltd.

Terminal restriction fragment length polymorphism(T-RFLP) analysisCyanobacterial community profiles of five lesionsthat visually changed from CP to BBD (lesions weresampled from 24 August to 22 September, 7 Octoberto 10 December, 24 August to 23 November, 25October to 10 December and 27 January to 15February, respectively) were examined by T-RFLP.Cyanobacterial 16S rRNA genes were PCR amplifiedusing the same reaction mixture and conditions asoutlined for the cyanobacterial clone libraries,except that 0.1 mM CYA106F with the WellRED D4fluorescent label at the 50 end (Sigma-Aldrich) and

Bacterial transition in the development of BBDY Sato et al

206

The ISME Journal

0.1 mM CYA781R were used as the primers. LabeledPCR products were quantified with a NanoDropSpectrophotometer (Thermo Fisher Scientific) and75 ng of each product was digested with therestriction enzyme RsaI (New England Biolabs).Digested fragments were purified with the DyeEx2.0 Spin Kit (Qiagen). Lengths of terminal restrictionfragments (T-RFs) were determined and visualizedusing the CEQ8800 Genetic Analysis System(Beckman Coulter, Fullerton, CA, USA) with a sizestandard (600bp, Beckman Coulter) following themanufacturer’s instructions. Resulting T-RFLP pro-files were compared with ones obtained fromcyanobacterial clones representing each of the OTUsin the same manner.

Phylogenic analysis of clone sequencesRetrieved cyanobacterial and bacterial 16S rRNAgene clone sequences were visualized and vectorsequences were removed with the sequence analysispackage Sequencher (Gene Codes Corporation, AnnArbor, MI, USA). Sequences were imported into theARB software package (http://www.arb-home.de)(Ludwig et al., 2004), and aligned against the Green-genes database (http://greengenes.lbl.gov) (DeSantiset al., 2006), followed by a manual correction ofthe alignment when necessary. Additional referencesequences not available in the Greengenes databaseat the time of analysis were identified from BLASTsearches (Altschul et al., 1997) of sequences re-trieved in this study, imported and aligned usingARB. Phylogenetic trees were constructed using theneighbor-joining (Jukes–Cantor correction) (Saitouand Nei, 1987) algorithms implemented in ARB. Therobustness of the inferred tree topologies wasevaluated after 1 000 bootstrap replicates of theneighbor-joining data. Partial sequences (B700 bp)were inserted into the tree without changing the treetopology by using the ARB parsimony interactivemethod.

Nucleotide sequence accession numbersAll unique sequences retrieved from this study havebeen deposited in the GenBank database underaccession numbers GQ204788–GQ204978.

Results

Field observations of CP and BBDCyanobacterial patch lesions on laminar and en-crusting colonies of Montipora characteristicallyappeared either as small (1–5 cm in diameter), palegreen/brown, semicircular lesions at the edge ofcolonies or as round patches surrounded by appar-ently healthy coral tissue in the middle of colonies(Figure 1a: CP). CP lesions typically expandedradially across coral tissues and subsequentlydarkened at their edges (intermediate stage;Figure 1b: IM), leaving exposed skeleton that

gradually became covered with algal turf behindthe lesion front. Darkened edges of CP lesionsvisually changed into the dark microbial matcharacteristic of BBD, which also migrated acrosscoral colonies, killing tissue and leaving freshlyexposed white skeleton (Figure 1c: BBD). From atotal of 262 CP lesions recorded on Montiporacolonies monitored between September 2006 andJanuary 2009, 18.7% developed into the visuallycharacteristic BBD lesions. The remaining CPlesions disappeared leaving exposed coral skeletoncovered with turf algae without an actively progres-sing microbial mat. These CP-derived BBD lesionsaccounted for 18.6% of a total of 263 BBD cases thatwere observed within this Montipora assemblageduring the same period. Sources of infection for theremaining BBD lesions were unknown, except for1.5% of infections that were apparently caused bydirect contact with a BBD lesion on a neighboringcolony.

The 2.3-year monitoring program documentedthat maximum prevalence of CP occurred betweenNovember and December each year, followed by apeak in BBD prevalence 40–50 days later (Figure 2).Specifically, characteristic BBD lesions developedfrom tagged CP lesions 62±5 days (mean±s.e.,n¼ 49 lesions) after the corresponding CP was firstrecorded.

Annual mean (±s.e.) rates of linear progressionfor lesion fronts were 0.47 ± 0.03 mm per day for CP(n¼ 55 lesions) and 2.38±0.12 mm per day for BBD(n¼ 65 lesions). Linear progression rates of BBDwere significantly higher than those of CP lesions,regardless of the season (two-way ANOVA,F¼ 160.5, df¼ 1, Po0.001: Figure 3). CP and BBDlesions both progressed more rapidly in summerthan in winter.

BBDCP

Pre

vale

nce

(%)

15

10

5

0

Month

3 3

3

3

366

6

666

636

666

S N M M J S N JJ M M J SO D F A J A O D F A J AA N JO D F'06 '09'07 '08

Figure 2 Temporal patterns in the prevalence of CP and BBD in aMontipora species assemblage on the east coast of Pelorus Island,central Great Barrier Reef. Plots and error bars represent meanprevalence ± s.e. Numbers above or below plots indicate samplesize (quadrats; note that three quadrats were established in May2007 and were not accessible in April 2008). Abbreviations: BBD,black band disease; CP, cyanobacterial patch.

Bacterial transition in the development of BBDY Sato et al

207

The ISME Journal

Microscopic observations of CP and BBDHistological examination revealed that microbialmats of both CP and BBD were dominated bycyanobacterial filaments, although the dominantcyanobacteria within the respective microbial matswere morphologically distinct (Figures 4a and b).Cyanobacterial filaments within CP lesions weretypically twofold greater in diameter (9.0–9.2mm)than filaments in the BBD mat (4.0–4.2 mm). BothCP- and BBD-derived cyanobacteria showed glidingmotility. At the interface between the CP lesion andapparently healthy tissue, free body wall coraltissues (for example, tentacles, oral disks andcoenosarc) were degraded beneath the CP bacterialmat, however, aboral structures (calicoblastic layerand lower mesentries) appeared intact (Figure 4c).In contrast, tissue structure beneath the BBD matat the interface area was severely disorganized fromthe oral to the aboral surface (Figure 4d). Trichomespenetrating into CP-infected coral tissue were observedat the base of apparently intact epidermis (Figure 4e),whereas BBD cyanobacteria were observed throughoutthe necrotic tissue mass including the gastrodermis(Figure 4f).

Molecular comparisons of cyanobacteria betweenCP and BBDCyanobacterial-specific 16S rRNA gene clonelibraries derived from microbial mats were distinctfor CP and BBD lesions (Table 1). CP clone libraries(n¼ 2 lesions) were dominated by sequences (desig-nated as CL35–CP–OTU1) closely affiliated with aBlennothrix sp. In contrast, BBD libraries (n¼ 2lesions) were dominated by sequences (designatedas CL36–BBD–OTU1) closely related to ones thathave been retrieved from BBD lesions on Red Seacorals (BB1S16SI-18). This BBD dominant clone type,CL36–BBD–OTU1, also appeared in the CP libraries,although it represented only 1.5% of each library.Similarly, the CP dominant clone type, CL35–CP–OTU1, was retrieved from the BBD lesion of colony1, representing 11.6% of the clone library. Ambientseawater collected above the diseased corals showed

a higher diversity of cyanobacterial ribotypes,although they were distinct from ribotypes retrievedfrom CP and BBD lesions, with only one sequence(CL37–SW–OTU1), related to an uncultured Syne-chococcus species MC21, shared with one CP library(colony 1).

Terminal restriction fragment length polymorph-ism (T-RFLP) profiles of amplified cyanobacterial16S rRNA genes illustrated successional changes indominating T-RFs from 343/344 to 149 nucleotidelength fragments as CP changed into BBD (Figure 5).Five replicate coral colonies showed this iden-tical T-RF pattern (including colonies 1–3, data notshown). T-RFLP analyses of representative cyano-bacterial 16S rRNA gene clones confirmed that the343/344 and 149 peaks corresponded to CL35–CP–OTU1 and CL36–BBD–OTU1, respectively, whichalso matched with theoretical lengths of T-RFs ofCL35–CP–OTU1 and CL36–BBD–OTU1 digested withthe RsaI restriction enzyme.

Phylogenetic analysis of 16S rRNA gene partialsequences indicated that the dominant cyanobacter-ial sequences retrieved from the CP and BBD lesionswere positioned in two distant groups in compar-ison with other reference cyanobacteria (Figure 6).CL35–CP–OTU1 formed a cluster related to Tricho-desmium species, although it was more closelyaffiliated with a group including a Blennothrixspecies, an Oscillatoria sp. PAB-21 and PNG-50retrieved from a BBD lesion on corals in Papua NewGuinea. CL36–BBD–OTU1 and closely related BBDcyanobacterial sequences retrieved from corals inthe Red Sea (BB1S16SI-18), Palau (BMS1) and theCaribbean (128–56 and CD1C11) formed a distinctgroup.

Bacterial shifts during the transition from CP to BBDThe composition of bacterial 16S rRNA gene clonelibraries derived from CP and BBD lesions samp-led from the same colony showed distinctive phy-logenetic shifts in higher bacterial taxonomicgroups as the lesions transitioned from CP to BBD(n¼ 3 colonies, Supplementary Tables S1). In gen-eral, Alphaproteobacteria-affiliated sequences weredominant in CP clone libraries (56.8–63.6% oflibraries), followed by Gammaproteobacteria-relatedsequences (26.1–33.0% of libraries). In contrast,BBD libraries were dominated by sequences affi-liated to Gammaproteobacteria (23.0–57.0% oflibraries). Alphaproteobacteria represented 30.1%of the BBD library of colony 1 but were only a minorconstituent (5.5–11.5%) of the other two replicatelibraries. Cyanobacteria were also more commonlyretrieved from BBD libraries than CP libraries,representing 57.7% of clones in the BBD library ofcolony 3. Deltaproteobacteria-affiliated sequences,including Desulfobacter sp. and Desulfovibriosp., were retrieved from two BBD libraries, althoughone Desulfovibrio sp.-affiliated sequence wasalso retrieved from one CP library (colony 2).

BBD

CP

Summer Winter0

1

2

3

Line

ar p

rogr

essi

on (

mm

d-1

)

(9)

(44)

(17)

(7)

Figure 3 Linear progression rates of CP and BBD on Montiporaspecies during summer and winter. Columns and bars representmeans ± s.e. Numbers in brackets indicate sample size (lesions).Abbreviations: BBD, black band disease; CP, cyanobacterial patch.

Bacterial transition in the development of BBDY Sato et al

208

The ISME Journal

Epsilonproteobacteria- and Deferribacteres-relatedsequences were retrieved only from BBD clonelibraries.

The bacterial communities of disease lesionsshowed successional changes during the transitionfrom CP to BBD, as illustrated by clone librariesconstructed for a lesion on colony 3 that wasrepeatedly sampled (Figure 7 and SupplementaryTable S2). First, the libraries from the early stages ofthe CP lesion (CP1 and CP2) contained a limitednumber of cyanobacterial-affiliated sequences. Evenat the intermediate stage (IM), the relative number ofcyanobacterial sequences represented only 7.3% ofthe clone library, although interestingly, these cyano-bacterial sequences shifted from those dominatingthe early CP libraries (CP1 and CP2) to sequencesclosely related to previously retrieved BBD cyano-bacterial sequences (Supplementary Table S2). Whenthe characteristic BBD signs were first visuallyobserved (BBD1), the relative number of cyano-bacterial sequences increased to 42.6% and increased

further as the lesion developed into a rapidly expand-ing BBD lesion (57.7% of BBD2 library). Secondly,coinciding with this increase in relative number ofcyanobacterial-affiliated sequences, Epsilonproteo-bacteria-affiliated sequences were retrieved in theBBD1 and BBD2 libraries, although these repre-sented only a small relative percentage of cloneswithin the libraries (1.1–2.6%), whereas the relativeabundance of alphaproteobacterial sequences de-clined. Thirdly, Desulfovibrio spp.-affiliated se-quences first appeared in the IM library andsubsequently formed small percentages (1.3–4.6%)of the BBD1 and BBD2 libraries. The identity ofsequences retrieved from all clone libraries and theirclosest affiliations are compared in SupplementaryTable S3. Supplementary Table S3 also illustratesbacterial sequences retrieved from a healthy coraltissue sample (colony 4—HC) and highlights thedistinctiveness of this library, which shares only afew individual sequences with CP- and BBD-derivedclone libraries.

CF

CF

EP

EP

EP

GAGA

ZXCF

CF CF

CF

ZX

Figure 4 Microscopic images showing typical appearance of cyanobacterial filaments derived from: (a) CP and (b) BBD. Histologicalappearance of the interface between a microbial mat and tissues of the coral, Montipora hispida, for: (c, e) CP; and (d, f) BBD.Abbreviations: BBD, black band disease; CF, cyanobacterial filament; CP, cyanobacterial patch; EP, epidermis; GA, gastrodermis; ZX,zooxanthella.

Bacterial transition in the development of BBDY Sato et al

209

The ISME Journal

Discussion

Field monitoring throughout seasonal outbreaks ofBBD within an assemblage of Montipora species inthe central GBR confirmed that, in some cases,characteristic BBD develops from visually distinc-tive, cyanobacterium-infected lesions, herein named‘cyanobacterial patch(es) (CP)’. The 40- to 50-day earlierpeak in the prevalence of CP compared with the peakin BBD prevalence, which roughly coincides with theaverage length of time for individually monitored CPlesions to develop into BBD, provides corroborativeevidence that CP represents an early successional stagein the development of some BBD cases.

Histological and microbiological investigationsindicated that CP and BBD lesions comprisedmorphologically and phylogenetically distinct cya-nobacterial assemblages and that this shift occurredthrough successional changes in microbial assem-blages, which correlated with field observations ofdisease signs. Cyanobacterial 16S rRNA gene clonelibraries derived from CP lesions were dominatedby Blennothrix sp.-affiliated sequences (CL35–CP–OTU1), whereas libraries from BBD lesions weredominated by a phylogenetically distinct sequence(CL36–BBD–OTU1), which formed a tight clusterwith sequences retrieved from BBD mats in theRed Sea (Barneah et al., 2007), Palau (SussmanT

able

1C

om

posi

tion

san

did

en

tity

of

part

ial

16S

rRN

Agen

ese

qu

en

ces

am

pli

fied

wit

hcyan

obac

teri

al-

specif

icp

rim

ers

(CY

A106F

/CY

A781R

)in

clo

ne

libra

ries

deri

ved

from

mic

robia

lm

ats

of

cyan

obacte

rial

patc

han

dbla

ckban

dd

isease

on

two

Mon

tip

ora

his

pid

acolo

nie

san

dam

bie

nt

seaw

ate

rabove

the

sam

ple

dcolo

nie

s

Sequ

ence

IDA

ccess

ion

nu

mber

Clo

sest

rela

tive

a(A

ccess

ion

no.)

Iden

tity

acolo

ny

1b

colo

ny

3

CP

(24

Au

g.’

07)c

BB

D(2

2S

ep

.’07)

CP

(7O

ct.

’07)

BB

D(1

0N

ov.

’07)

SW

(24

Au

g.’07)

CL

35–C

P–O

TU

1G

Q204788

Ble

nn

oth

rix

sp.

PN

G05–4

(EU

253968)

98%

61

860

CL

36–B

BD

–O

TU

1G

Q204789

Un

cu

ltu

red

BB

Dcyan

obacte

riu

mB

B1S

16S

I-18

(EF

433097)

99%

155

168

CL

37–S

W–O

TU

1G

Q204790

Un

cu

ltu

red

Syn

ech

ococcu

ssp

.M

C21

(DQ

903987)

99–100%

121

CL

35–C

P–O

TU

3G

Q204791

Un

cu

ltu

red

cyan

obacte

riu

m1D

P1-B

21

(EU

780374)

99%

1C

L28–B

BD

–O

TU

3G

Q204792

Tri

ch

od

esm

ium

ery

thra

eu

mIM

S101

(CP

000393)

94%

1C

L37–S

W–O

TU

2G

Q204793

Pro

ch

loro

coccu

sm

ari

nu

sM

IT9312

(CP

000111)

99%

49

CL

37–S

W–O

TU

3G

Q204794

Un

cu

ltu

red

Ple

cto

nem

asp

.C

8(D

Q072876)

96%

2C

L37–S

W–O

TU

4G

Q204795

Syn

ech

ococcu

ssp

.R

CC

307

(CT

978603)

99%

2C

L37–S

W–O

TU

5G

Q204796

Un

cu

ltu

red

cyan

obacte

riu

mA

822

(EU

283559)

98%

1N

on

cyan

obact

eri

al

sequ

en

ces

——

55

58

Su

m68

69

67

68

83

Abbre

via

tion

s:B

BD

,bla

ck

ban

dd

isease

;C

P,cyan

obacte

rial

patc

hes;

SW

,se

aw

ate

rsa

mp

le.

aS

equ

en

ces

were

ali

gn

ed

toth

eclo

sest

rela

tive

over

700

bp

usi

ng

BL

AS

T(A

ltsc

hu

let

al.

,1997).

Th

esi

mil

ari

tyw

as

calc

ula

ted

wit

hgap

sn

ot

taken

into

accou

nt.

bL

abeli

ng

of

colo

nie

scorr

esp

on

ds

wit

hM

ate

rials

an

dm

eth

od

s.cD

ate

sre

pre

sen

tw

hen

each

sam

ple

was

coll

ecte

d.

343/344

342/343

149

149

342

0 100 200 300 400 500 600

Fragment length (nt)

Rel

ativ

e flu

ores

cenc

e in

tens

ity

(CP)

(IM)

(BBD)

Figure 5 Results of terminal restriction fragment length poly-morphism (T-RFLP) analyses of microbial communities asso-ciated with lesions on the coral, Montipora hispida (colony 3),showing cyanobacterial communities from: (a) CP (7 October2007), (b) an intermediate stage between CP and BBD (25 October2007) and (c) BBD (10 November 2007). Approximately 700-bppartial sequences of 16S ribosomal RNA (rRNA) genes were PCRamplified with cyanobacterium-targeting primers and digestedwith the RsaI restriction enzyme. Numbers above peaks indicateexact sizes of T-RFs. Abbreviations: BBD, black band disease; CP,cyanobacterial patch; IM, intermediate stage between CP andBBD.

Bacterial transition in the development of BBDY Sato et al

210

The ISME Journal

et al., 2006) and the Caribbean (Cooney et al., 2002;Frias-Lopez et al., 2003). T-RFLP analyses ofreplicate samples (n¼ 5 colonies) all supported this

successional change in dominant T-RFs, with peakscorresponding to CL35–CP–OTU1 in CP and CL36–BBD–OTU1 in BBD samples. Cyanobacterium-af-filiated sequences in 16S rRNA gene clone librariesconstructed with the bacterial-specific 63F/1387Rprimers also support distinct CP and BBD cyano-bacterial assemblages: Red Sea BBD cyanobacte-rium (BB1S16SI-18)-affiliated sequences dominatedcyanobacterial portions of BBD libraries, whereassequences affiliated with an Oscillatoria sp. PAB-21appeared in CP libraries, which are phylogeneticallyclose to Blennothrix sp. PNG05–4 (98% sequenceidentity).

Phylogenetically, the dominant cyanobacterialsequence derived from CP lesions (CL35–CP–OTU1) clusters with Trichodesmium spp. seq-uences and is closely related to a Blennothrix(¼Hydrocoleum) sp.-affiliated sequence. Hydroco-leum spp. are the most common mat-formingbenthic cyanobacteria in tropical oceans (Abedet al., 2003), and they contribute the major compo-nent of nitrogen fixation in tropic lagoons (Charpyet al., 2007). If this cyanobacterium within the CPmat functions similarly, it may have an importantrole in bacterial succession from CP to BBD byincreasing local concentrations of fixed nitrogen andcreating new niches for other bacteria utilizing theseproducts. Hydrocoleum spp. are also morphologi-cally and phylogenetically close to planktonicspecies of Trichodesmium, suggesting a commonevolutionary origin (Abed et al., 2006). Interestingly,Trichodesmium spp.-related 16S rRNA gene se-quences have been retrieved previously from BBDbacterial communities as dominant cyanobacterialsequences, such as a Trichodesmium tenue-affilia-ted (93% sequence identity) sequence from theCaribbean (Frias-Lopez et al., 2002) and a sequencedesignated as PNG-50 from the Indo-Pacific (Frias-Lopez et al., 2003). Together, these results suggestthat the ‘Trichodesmium/Hydrocoleum cluster’(Abed et al., 2006) includes benthic cyanobacterialspecies that dominate pathogenic bacterial mats onlive coral colonies such as BBD and CP. Sequencesaffiliated with planktonic species recovered from aseawater clone library, such as Trichodesmiumerythraeum IMS101 and Synechococcus sp. MC21,also occurred in CP and BBD clone libraries.However, as these sequences constitute only minorcomponents in the CP and BBD clone libraries, itis unclear whether they are contaminants fromthe ambient water column or cyanobacteria play-ing specific roles in CP and BBD microbial com-munities.

Linear progression rates of BBD lesions werefivefold higher than those of CP lesions, suggestingthat the BBD bacterial community is more virulentthan the CP community. Although trichomes of bothCP- and BBD-dominating cyanobacterial morpho-types showed gliding motility and penetration intocoral tissues, BBD cyanobacteria were observed atdeeper histological layers within coral polyps than

Unc BB1S16SI-18 Red Sea BBD (EF433097)CL36-BBD-OTU1 (GQ204789) BMS1 Palau BBD (AY839640)Unc 128-56 Caribbean BBD (AF473936)Unc CD1C11 Caribbean BBD (AY038527)

Oscillatoria spongeliae (AY615504)Microcoleus chthonoplates (X70770)

Geitlerinema sp. BBD (DQ151461)Unc Limnothrix sp. EC2 (DQ889938)

Prochlorococcus marinus (CP000111)Unc Synechococcus sp. MC21 (DQ903987)

Synechococcus sp. RCC307 (CT978603)Leptolyngbya antarctica (AY493589)

Unc Plectonema sp. C8 (DQ072876)Trichodesmium tenue (AF013029)Trichodesmium contortum (AF013028)Trichodesmium erythraeum (CP000393)Trichodesmium hildebrandtii (AF091322)

Trichodesmium thiebautii (AF013027)Oscillatoria sp. PAB-21 (EU253967)Unc PNG-50 Indo-Pacific BBD (AY123040)

CL35-CP-OTU1 (GQ204788)

Blennothrix sp. PNG05-4 (EU253968)Oscillatoria sancta (AB039015)

Planktothrix agardhii (AB074507)0.05

Unc 216-32 Caribbean BBD (DQ446127)

Figure 6 Phylogenetic neighbor-joining tree of approximately700-bp partial 16S ribosomal RNA (rRNA) gene sequences ofcyanobacteria derived from cyanobacterial patches (CP) (CL35–CP–OTU1) and black band disease (BBD) (CL36–BBD–OTU1)with other reference sequences including cyanobacteria previouslyretrieved from BBD samples (labeled with ‘BBD’). Bar indicates0.05 change per base position.

100

80

60

40

20

0

Com

posi

tion

of c

lone

s (%

)

Disease stagesCP1 CP2 IM BBD1 BBD2

α-Proteobacteria

γ-Proteobacteria

δ-Proteobacteria

ε-Proteobacteria

Bacteroidetes

Firmicutes

Cyanobacteria

(88) (95) (82) (94) (78)

Figure 7 Comparison of clone library compositions of bacterial16S ribosomal RNA (rRNA) genes amplified with bacterial-specificprimers (63F and 1378R) throughout the transition from cyano-bacterial patch (CP1, 25 August 2007; CP2, 07 October 2007), inter-mediate stage (IM, 25 October 2007) and BBD (BBD1, 10 November2007; BBD2, 23 November 2007) from the identical lesion on aMontipora hispida colony (colony 3). Numbers in brackets abovecolumns indicate sizes of clone libraries (sequences).

Bacterial transition in the development of BBDY Sato et al

211

The ISME Journal

CP cyanobacteria, suggesting their greater potentialfor mechanical and/or chemical destruction ofcoral tissues. However, it is unclear whether this isbecause of a greater ability of the BBD cyanobacteriato penetrate into coral tissue or because the coraltissue was already degraded by other virulencefactors associated with the BBD bacterial commu-nity. Higher biomass of cyanobacteria in the BBDmat, suggested by the higher representation ofcyanobacterial-affiliated sequences within BBDclone libraries, is also possibly an important factorthat promotes faster progression of BBD lesionsthrough formation of a dense microbial communitythat allows development of anoxic zones andconsequently substantial diurnal changes betweenoxic and anoxic conditions (Carlton and Richard-son, 1995). Darkening in the color of lesions at theonset of characteristic BBD signs may be attributedto an increase in cyanobacterial densities. Further-more, cyanobacterial toxins have been suggested tohave a role in the pathogenicity of BBD (Richardsonet al., 2007). The toxicity of cyanobacterial bloomsof T. erythraeum and Trichodesmium thiebautii,phylogenetically close to CL35–CP–OTU1, has beenreported from the GBR and Caribbean waters,respectively (Hawser and Codd, 1991; Endeanet al., 1993). However, the presence of toxinsproduced by cyanobacterial strains dominating CPand BBD lesions cultured from the study site wasnot detected (unpublished results), although furtherassessment of the role of toxins in the etiology of CPand BBD is required.

Sequences involved in sulfur cycling appearedduring successional changes in bacterial commu-nities from CP to BBD, as illustrated by bacterial 16SrRNA gene clone libraries. In addition to anoxicconditions, the presence of highly concentratedsulfide at the base of the BBD mat has been shownto cause coral tissue necrosis (Richardson et al.,1997) and there is clear evidence that Desulfovibriospp. and other sulfate-reducing bacteria occupymicroenvironmental niches within the BBD mat(Cooney et al., 2002; Frias-Lopez et al., 2002; Sekaret al., 2006, 2008; Viehman et al., 2006; Barneahet al., 2007). This study retrieved sequencesaffiliated with Deltaproteobacteria, including De-sulfobacter sp. and Desulfovibrio sp. from BBD clonelibraries. One Desulfovibrio sp.-affiliated sequencewas detected in one CP library, suggesting that theseorganisms may already be present in the CP bacterialmat. Interestingly, Epsilonproteobacteria- andDeferribacteres-affiliated sequences were retrievedonly from BBD clone libraries. Epsilonproteobacterialsequences retrieved from BBD libraries wereaffiliated with Candidatus Arcobacter sulfidicus(94–97% sequence identity), a marine sulfide-oxi-dizing autotrophic bacterium that produces filamen-tous sulfur in sulfide-oxygen gradients (Wirsenet al., 2002). The Deferribacteres bacteria includeiron-, manganese- and nitrogen-reducers, found inanaerobic microenvironments (Greene et al., 1997;

Myhr and Torsvik, 2000). The sulfur cycle is oftentied to other elemental cycles such as nitrogen,iron, phosphorus and carbon within a micro-environment of highly diverse bacterial commu-nities (Sievert et al., 2007). Thus, the presence ofthese Epsilonproteobacteria and Deferribacteresspecies, as well as Gammaproteobacteria andDeltaproteobacteria, within BBD libraries likelyindicates that BBD lesions possess a more dynamicsulfur cycle with higher concentrations of toxicsulfide than CP lesions. Combined with highercyanobacterial biomass, these bacterial elementcycles may promote further anoxic sulfide richmicroenvironments and hence faster lesion progres-sion of BBD compared with CP. As the BBD bandforms, diurnal fluctuations of sulfide and oxygenconcentrations along the vertically stratified micro-environment are expected to occur (Carlton andRichardson, 1995). Under these rapidly changingmicroenvironmental conditions, bacteria that canwithstand such extreme fluctuations possiblybecome dominant in the microbial community(Sekar et al., 2008). To verify microenvironmen-tal regimes within the changing microbial mat fromCP to BBD, isolation and physiological characteri-zation of CP- and BBD-dominating cyanobacterialcultures along with microsensor studies are cur-rently underway.

Sequences affiliated with bacteria involved insulfur metabolism were, however, potentially under-represented in the clone library analyses. Thebacterial-specific primers used in this study (63F/1387R) match only 28% of Desulfovibrio spp.-, ando1% of Epsilonproteobacteria-affiliated sequencesthat are available in public databases (assessed byprimer comparison to the Ribosomal DatabaseProject II). Sulfide-oxidizing Beggiatoa spp. havebeen reported as constituents of the BBD microbialmat (Richardson, 1996; Viehman and Richardson,2002; Sekar et al., 2008), although ribotypesaffiliated with Beggiatoa spp. were not retrievedfrom CP or BBD clone libraries in this study. Thismay be due to primer bias or limitations, bias inDNA preparation (Sekar et al., 2009), and/or becausesequences of Beggiatoa strains specifically asso-ciated with BBD have not been deposited in publicdatabases (Frias-Lopez et al., 2002). Nonetheless, allsamples in this study were treated similarly andtherefore major shifts in bacterial ribotypes aredeemed to reflect the actual transition in bacterialcompositions from CP to BBD. Further investigationof bacterial communities using probes and primerstargeting specific bacterial groups and their func-tional genes (for example, Barneah et al., 2007) willprovide further insight into the development of theBBD bacterial consortium and roles of key membersin the transition from CP to BBD.

Approximately, 19% of CP lesions developed intoBBD, however, it is unclear what factors governwhether a CP lesion develops into BBD or loses theactive microbial front. Overall, bacterial ribotype

Bacterial transition in the development of BBDY Sato et al

212

The ISME Journal

composition derived from a CP lesion that did notdevelop into BBD was similar to that of CP lesionswhich did develop into BBD (Y Sato, 2008,unpublished data). Further investigations of intra-colony variation in bacterial members within CPlesions and of host immune responses are requiredto identify factors driving the development of BBDfrom CP lesions. Although BBD has been reportedfrom throughout the GBR on multiple host species(Page and Willis, 2006), CP lesions have beenobserved only on Montipora species in the cen-tral, inshore GBR. Therefore species- and location-specific disease mechanisms leading to the onset ofBBD may exist. Earlier reports have shown that BBDcan be triggered from white band disease (Antonius,1985) on scleractinian corals despite the absence ofa visible microbial mat in this case. This studyindicates that microbial lesions with a distinctbacterial community such as CP may develop intothe characteristic BBD through successional changesin bacterial communities, but this may not be theonly or even the major mode of onset of BBD.

In summary, BBD lesions showed faster progres-sion and hence higher virulence than CP lesions,potentially caused by distinct bacterial assemblagesin the BBD microbial mat, such as the dominantcyanobacterial and sulfur-cycle-associated species.Formation of the characteristic BBD band may befacilitated by the bacterial community in CP lesionsthrough alteration of microenvironmental condi-tions and creation of microecological niches. Differ-ences in bacterial composition between CP and BBDalso suggest that bacterial pathogen(s) that triggerthe BBD infection may become less abundant duringsubsequent successional changes. This may be onereason why identifying the primary causative agent(s)of BBD has been difficult. Further investigationsof key functional genes and their expressions atdifferent disease stages in the CP–BBD systemwill provide valuable information to elucidate whatrole each bacterial member performs within themicrobial communities and thus, which bacteriaare key players in the development of BBDpathogenesis.

Conflict of interest

The authors declare no conflict of interest.

Acknowledgements

This study was funded by AIMS@JCU, the ARC Centre ofExcellence for Coral Reef Studies, and the GEF DiseaseWorking Group in the Coral Reef Targeted Research andCapacity Building for Management Program. We thank thestaff of Orpheus Island Research Station (JCU) for logisticsupport, Tim Simmonds for graphical support andnumerous volunteers for field assistance.

References

Abed RMM, Golubic S, Garcia-Pichel F, Camoin GF,Sprachta S. (2003). Characterization of microbialite-forming cyanobacteria in a tropical lagoon: TikehauAtoll, Tuamotu, French Polynesia. J Phycol 39:862–873.

Abed RMM, Palinska KA, Camoin G, Golubic S. (2006).Common evolutionary origin of planktonic andbenthic nitrogen-fixing oscillatoriacean cyanobacteriafrom tropical oceans. FEMS Microbiol Lett 260:171–177.

Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z,Miller W et al. (1997). Gapped BLAST and PSI-BLAST:a new generation of protein database search programs.Nucleic Acids Res 25: 3389–3402.

Antonius A. (1981). The ‘band’ diseases in coral reefs.Proc Fourth Int Coral Reef Symp 2: 7–14.

Antonius A. (1985). Black band disease infection experi-ments on hexacorals and octocorals. Proc Fifth IntCoral Reef Cong 6: 155–160.

Barneah O, Ben-Dov E, Kramarsky-Winter E, Kushmaro A.(2007). Characterization of black band disease in RedSea stony corals. Environ Microbiol 9: 1995–2006.

Borger JL, Steiner SCC. (2005). The spatial and temporaldynamics of coral diseases in Dominica, West Indies.Bull Mar Sci 77: 137–154.

Bruckner AW, Bruckner RJ. (1997). The persistence ofblack-band disease in Jamaica: impact on communitystructure. Proc Eighth Int Coral Reef Symp 1: 601–606.

Bruckner AW, Bruckner RJ, Williams EH. (1997). Spread ofa black-band disease epizootic through the coralreef system in St Ann’s Bay, Jamaica. Bull Mar Sci61: 919–928.

Carlton RG, Richardson LL. (1995). Oxygen and sulfidedynamics in a horizontally migrating cyanobacterialmat — Black band disease of corals. FEMS MicrobiolEcol 18: 155–162.

Charpy L, Alliod R, Rodier M, Golubic S. (2007). Benthicnitrogen fixation in the SW New Caledonia lagoon.Aquat Microb Ecol 47: 73–81.

Cooney RP, Pantos O, Tissier MDAL, Barer MR, O0DonnellAG, Bythell JC. (2002). Characterization of the bacter-ial consortium associated with black band disease incoral using molecular microbiological techniques.Environ Microbiol 4: 401–413.

DeSantis TZ, Hugenholtz P, Larsen N, Rojas M, Brodie EL,Keller K et al. (2006). Greengenes, a chimera-checked16S rRNA gene database and workbench compatiblewith ARB. Appl Environ Microbiol 72: 5069–5072.

Endean R, Monks SA, Griffith JK, Llewellyn LE. (1993).Apparent relationships between toxins elaborated bythe cyanobacterium Trichodesmium erythraeum andthose present in the flesh of the narrow-barred spanishmackerel Scomberomorus commersoni. Toxicon 31:1155–1165.

Frias-Lopez J, Bonheyo GT, Jin QS, Fouke BW. (2003).Cyanobacteria associated with coral black band dis-ease in Caribbean and Indo-Pacific Reefs. ApplEnviron Microbiol 69: 2409–2413.

Frias-Lopez J, Zerkle AL, Bonheyo GT, Fouke BW. (2002).Partitioning of bacterial communities between sea-water and healthy, black band diseased, and deadcoral surfaces. Appl Environ Microbiol 68: 2214–2228.

Goreau TJ, Cervino J, Goreau M, Hayes R, Hayes M,Richardson L et al. (1998). Rapid spread of diseases inCaribbean coral reefs. Rev Biol Trop 46: 157–171.

Bacterial transition in the development of BBDY Sato et al

213

The ISME Journal

Green EP, Bruckner AW. (2000). The significance of coraldisease epizootiology for coral reef conservation. BiolConserv 96: 347–361.

Greene AC, Patel BKC, Sheehy AJ. (1997). Deferribacterthermophilus gen nov, sp nov, a novel thermophilicmanganese- and iron-reducing bacterium isolatedfrom a petroleum reservoir. Int J Syst Bacteriol 47:505–509.

Hawser SP, Codd GA. (1991). The toxicity of Trichodes-mium blooms from Caribbean waters. In: Carpenter EJ,Capone DG, Rueter JG (eds). Marine Pelagic Cyano-bacteria: Trichodesmium and Other Diazotrophs.Kluwer Academic Publishers: Dordrecht. pp 319–327.

Kaczmarsky LT. (2006). Coral disease dynamics in thecentral Philippines. Dis Aquat Org 69: 9–21.

Klaus JS, Janse I, Heikoop JM, Sanford RA, Fouke BW.(2007). Coral microbial communities, zooxanthellaeand mucus along gradients of seawater depth andcoastal pollution. Environ Microbiol 9: 1291–1305.

Kuta KG, Richardson LL. (2002). Ecological aspects ofblack band disease of corals: relationships betweendisease incidence and environmental factors. CoralReefs 21: 393–398.

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

Marchesi JR, Sato T, Weightman AJ, Martin TA, Fry JC,Hiom SJ et al. (1998). Design and evaluation of usefulbacterium-specific PCR primers that amplify genescoding for bacterial 16S rRNA. Appl Environ Microbiol64: 795–799.

Myers JL, Richardson LL. (2009). Adaptation of cyano-bacteria to the sulfide-rich microenvironment of blackband disease of coral. FEMS Microbiol Ecol 67:242–251.

Myers JL, Sekar R, Richardson LL. (2007). Moleculardetection and ecological significance of the cyanobac-terial genera Geitlerinema and Leptolyngbya in blackband disease of corals. Appl Environ Microbiol 73:5173–5182.

Myhr S, Torsvik T. (2000). Denitrovibrio acetiphilus, anovel genus and species of dissimilatory nitrate-reducing bacterium isolated from an oil reservoirmodel column. Int J Syst Evol Microbiol 50:1611–1619.

Nubel U, Garcia-Pichel F, Muyzer G. (1997). PCR primersto amplify 16S rRNA genes from cyanobacteria. ApplEnviron Microbiol 63: 3327–3332.

Page C, Willis B. (2006). Distribution, host range and large-scale spatial variability in black band disease pre-valence on the Great Barrier Reef, Australia. Dis AquatOrg 69: 41–51.

Richardson LL. (1996). Horizontal and vertical migrationpatterns of Phormidium corallyticum and Beggiatoaspp. associated with black-band disease of corals.Microb Ecol 32: 323–335.

Richardson LL. (2004). Black band disease. In: RosenbergE, Loya Y (eds). Coral Health and Disease. Springer-Verlag: Heidelberg. pp 325–336.

Richardson LL, Kuta KG, Schnell S, Carlton RG. (1997).Ecology of the black band disease microbial consor-tium. Proc Eighth Int Coral Reef Symp 1: 597–600.

Richardson LL, Sekar R, Myers JL, Gantar M, Voss JD,Kaczmarsky L et al. (2007). The presence of thecyanobacterial toxin microcystin in black band diseaseof corals. FEMS Microbiol Lett 272: 182–187.

Rodriguez S, Croquer A. (2008). Dynamics of black banddisease in a Diploria strigosa population subjected toannual upwelling on the northeastern coast of Vene-zuela. Coral Reefs 27: 381–388.

Rohwer F, Seguritan V, Azam F, Knowlton N. (2002).Diversity and distribution of coral-associated bacteria.Mar Ecol Prog Ser 243: 1–10.

Saitou N, Nei M. (1987). The Neighbor-Joining Method—aNew Method for Reconstructing Phylogenetic Trees.Mol Biol Evol 4: 406–425.

Sato Y, Bourne DG, Willis BL. (2009). Dynamics of seasonaloutbreaks of black band disease in an assemblage ofMontipora species at Pelorus Island (Great Barrier Reef,Australia). Proc R Soc B Biol Sci 276: 2795–2803.

Schauer M, Massana R, Pedros-Alio C. (2000). Spatialdifferences in bacterioplankton composition along theCatalan coast (NW Mediterranean) assessed by mole-cular fingerprinting. FEMS Microbiol Ecol 33: 51–59.

Sekar R, Mills DK, Remily ER, Voss JD, Richardson LL.(2006). Microbial communities in the surface muco-polysaccharide layer and the black band microbial matof black band-diseased Siderastrea siderea. ApplEnviron Microbiol 72: 5963–5973.

Sekar R, Kaczmarsky LT, Richardson LL. (2008). Microbialcommunity composition of black band disease on thecoral host Siderastrea siderea from three regions of thewider Caribbean. Mar Ecol Prog Ser 362: 85–98.

Sekar R, Kaczmarsky LT, Richardson LL. (2009). Effect offreezing on PCR amplification of 16S rRNA genes frommicrobes associated with black band disease of corals.Appl Environ Microbiol 75: 2581–2584.

Sievert SM, Kiene RP, Schulz-Vogt HN. (2007). The sulfurcycle. Oceanography 20: 117–123.

Sussman M, Bourne DG, Willis BL. (2006). A singlecyanobacterial ribotype is associated with both redand black bands on diseased corals from Palau. DisAquat Org 69: 111–118.

Sutherland KP, Porter JW, Torres C. (2004). Disease andimmunity in Caribbean and Indo-Pacific zooxanthel-late corals. Mar Ecol Prog Ser 266: 273–302.

Viehman S, Mills DK, Meichel GW, Richardson LL. (2006).Culture and identification of Desulfovibrio spp. fromcorals infected by black band disease on Dominicanand Florida Keys reefs. Dis Aquat Org 69: 119–127.

Viehman TS, Richardson LL. (2002). Motility patterns ofBeggiatoa and Phormidium corallyticum in black banddisease. Proc Ninth Int Coral Reef Symp 2: 1251–1255.

Voss JD, Mills DK, Myers JL, Remily ER, Richardson LL.(2007). Black band disease microbial communityvariation on corals in three regions of the widerCaribbean. Microb Ecol 54: 730–739.

Voss JD, Richardson LL. (2006). Coral diseases near LeeStocking Island, Bahamas: patterns and potentialdrivers. Dis Aquat Org 69: 33–40.

Wirsen CO, Sievert SM, Cavanaugh CM, Molyneaux SJ,Ahmad A, Taylor LT et al. (2002). Characterization ofan autotrophic sulfide-oxidizing marine Arcobacter spthat produces filamentous sulfur. Appl EnvironMicrobiol 68: 316–325.

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

Bacterial transition in the development of BBDY Sato et al

214

The ISME Journal