microbial diversity within the water column of a larval rearing system for the ornate rock lobster (panulirus ornatus)

microbial diversity within the water column of a larval rearing system for the ornate rock lobster (panulirus ornatus) 作文 手机作文网 Microbial diversity within the water column of a larval rearing system for the ornate rock lobster (Panulirus ornatus) Microbial diversity within the water column of a larval rearingsystem for the ornate rock lobster (Panulirus ornatus)Matthew S. Paynea,b,?, Mike R. Halla, Raymond Bannistera,c, Lindsay Slyb,David G. BourneaaAustralian Institute of Marine Science, Tropical Aquaculture, PMB No 3, Townsville Mail Centre, QLD 4810, AustraliabSchool of Molecular and Microbial Sciences, University of Queensland, St Lucia, QLD 4067, AustraliacSchool of Marine Biology and Aquaculture, James Cook University, Townsville, QLD 4811, AustraliaReceived 30 January 2006; received in revised form 27 March 2006; accepted 2 April 2006AbstractThe ornate tropical rock lobster, Panulirus ornatus has substantial potential as an aquaculture species though disease outbreaksduring the animal’s extended larval lifecycle are major constraints for success. In order to effectively address such disease-relatedissues, an improved understanding of the composition and dynamics of the microbial communities in the larval rearing tanks isrequired. This study used flow cytometry and molecular microbial techniques (clone libraries and denaturing gradient gelelectrophoresis (DGGE)) to quantify and characterise the microbial community of the water column in the early stages(developmental stage I–II) of a P. ornatus larval rearing system. DGGE analysis of a 5000 L larval rearing trial demonstrated adynamic microbial community with distinct changes in the community structure after initial stocking (day 1 to day 2) and from day4 to day 5, after which the structure was relatively stable. Flow cytometry analysis of water samples taken over the duration of thetrial demonstrated a major increase in bacterial load leading up to and peaking on the first day of the initial larval moult (day 7),before markedly decreasing prior to when N50% of larvae moulted (day 9). A clone library of a day 10 water sample takenfollowing a mass larval mortality event reflected high microbial diversity confirmed by statistical analysis indices. Sequencesretrieved from both clone library and DGGE analyses were dominated by γ- and α-Proteobacteria affiliated organisms withadditional sequences affiliated with β- and ε-Proteobacteria, Bacteroidetes, Cytophagales and Chlamydiales groups. Vibrioaffiliated species were commonly retrieved in the clone library, though absent from DGGE analysis.Crown Copyright ? 2006 Published by Elsevier B.V. All rights reserved.Keywords: Panulirus ornatus phyllosoma; Bacterial diversity; 16S rRNA; DGGE; Clone library; Flow cytometry1. IntroductionThe global market demand for rock lobsters continuesto exceed supply from the wild and as such there is anincreasing interest in developing an aquaculture sector.Due to their value and importance in Australian wildfisheries, there has been particular emphasis on Jasusedwardsii ($177 million), J. verreauxi ($4.7 million) andPanuliruscygnus($305million)tobecomeestablishedasaquaculture species (ABARE, 2003). However, thesespecies do not necessarily possess favourable attributesfor culture. Although the wild fishery for the tropicalornate rock lobster (Panulirus ornatus), being 1 of 6tropical species in Australia, is currently valued at only$5 million (ABARE, 2003), this species has particularlyAquaculture 258 (2006) 80–90www.elsevier.com/locate/aqua-online?Corresponding author. Australian Institute of Marine Science,Tropical Aquaculture, PMB No 3, Townsville Mail Centre, QLD 4810,Australia. 作文 手机作文网 作文 手机作文网 Tel.: +61 7 4753 4139; fax: +61 7 4772 5852.E-mail addresses: m.payne@aims.gov.au (M.S. Payne),d.bourne@aims.gov.au (D.G. Bourne).0044-8486/$ - see front matter. Crown Copyright ? 2006 Published by Elsevier B.V. All rights reserved.doi:10.1016/j.aquaculture.2006.04.001 good traits as an aquaculture candidate. P. ornatus is thefastest growing of the lobsters of the family Palinuridaeand possesses one ofthe shortest known larval phases ofany rock lobster species, at approximately 160 days. Inaddition, the growth rate of P. ornatus under cultureconditionscanbeincreasedfurtherthroughthe manipula-tion of water temperature (Dennis et al., 1997; Lellis andRussell, 1990) and hormones (Juinio-Menezm andRuinata, 1996).The water column within larval rearing systems istypically a complex environment that may vary signifi-cantly from the larvae’s natural environment, both innutrient levels and microbial communities. Animalsreared in an aquaculture environment are subjected to avariety of stresses, including high stocking densities andsub-optimal water conditions. As a result, mech-anisms in culture animals are commonly adverselyaffected, and in an environment defense that is known to harbournumerous pathogens, this results in an increased suscep-tibility to disease (Takahashi et al., 1995). The commonpractice of using antibiotic and chemical treatments suchas formalin in procedures prior to and during larvalrearing, changes the water column microbial community.This is due to the ability of these treatments to removesome, but not all bacteria present in the water column,effectively selecting for certain strains.To date, there has been little research published on theoverall microbial community associated with wild orcultured rock lobster phyllosoma larvae and their en-vironments. The majority of current research has insteadfocused on the identification of pathogenic organismsfrom the phyllosoma themselves. This research has re-ported that Vibrio species are the most common pathogenamongst phyllosoma larvae (Webster et al., in press;Bourneetal.,2004;Digglesetal.,2000;Handlingeretal.,1999; Handlinger et al., 2000). In a previous study wereportedonthemicrobialdiversityofthecompartmentsofan entire P. ornatus larval rearing system, including thewater column, biofilm and phyllosoma (Bourne et al.,2004). This study reported the presence of a number ofgenera within the water column, including, Alcanivoraxsp.,Bacillussp.,Pseudoalteromonassp.andSulfitobactersp. In addition, Vibrio sp. were reported to be the domi-nant organism with Vibrio parahaemolyticus being thedominant species.To effectively manage the microbial communitywithin a larval rearing system, an understanding of themicrobial composition and dynamics within each com-partment (water column, phyllosoma, biofilm and livefeeds) is required to interpret how community changesmay affect phyllosoma health. The aim of this study wasto use a range of techniques to characterise the microbialcommunity of one compartment, the water column of anexperimental P. ornatus larval rearing system and de-termine how the microbial community changes through-out early stage larval rearing. The study focused on earlystage larval rearing, as this is a time that is frequentlyassociated with high larval mortalities in a number ofcultured crustaceans.2. Materials and methods2.1. Larval rearing technologyThe larval rearing process including broodstocksource, treatment and spawning, tank design, stockingmethods, feeding regimes and water recirculation andtreatment regimes was conducted as per the methods ofBourne et al. (2004). Moulting from phyllosoma stage1–2 (P1–P2) occurred from day 7–12, and a larval massmortality event occurred on day 10 as observed by alarge number of dead larvae ‘patches’ on the bottom ofthe larval rearing tank.2.2. Sample collectionTwo litres of water from the water column of the5000 L larval rearing tank was sampled daily in themorning, prior to the water exchange, using pre-sterilized equipment. 作文 手机作文网 作文 手机作文网 Samples were filtered through a100 μm filter to remove phyllosoma larvae and Artemia,with aliquots of collected water used for subsequentanalyses. Samples for flow cytometry analysis werecollected each day of the larval rearing run, up until theday of a larval mass mortality event. Samples for DGGEanalysis were collected each day up until five days past alarval mass mortality event. The water samples for clonelibrary analysis were collected on the day of a larvalmass mortality event.2.3. Flow cytometryTriplicate water samples (2 mL) were fixed with0.2 μm filtered electron microscopy grade 3% glutaral-dehyde (Electron Microscopy Sciences) in artificialseawater (17.55 gm NaCl; 0.75 gm KCl; 0.285 gmNa2SO4; 5.10 gm MgCl26H2O; 0.145 gm CaCl2) andstored at ?80 ?C until required. A 100 μL volume ofsample was added to 900 μL of 0.2 μM filtered sheathfluid(BectonDickinson)andanalysedasperthemethodsof Marie et al. (1999) using a FACScalibur (BectonDickinson)flowcytometer.Sampleswererunfor1minata flow rate of 12 μL min?1. Green fluorescence (fromsamplesstainedwithSYBRGreen)wascollectedthrough81M.S. Payne et al. / Aquaculture 258 (2006) 80–90 band pass interface filters at 530 nm. The three measuredparameters, green fluorescence(GFL),forward light scat-ter (FALS) and right-angle light scatter (RALS) -decadelogarithmicscale and analysed with the wererecordedandsortedinlistmodeona3 CellQuest Pro software pack-age (Becton Dickinson). Heterotrophic bacteria (HBAC)counts were recorded from water samples. Bacterial po-pulations were confirmed by running a positive controlcontaining known populations of a Vibrio sp. that hadbeenpreviouslyisolatedfromtheP.ornatuslarvalrearingsystem. All samples and controls were standardized with1 μm yellow/green fluorescent beads (Polysciences,Pennsylvania, USA).2.4. Extraction of genomic bacterial DNALarval rearing tank water samples (1 L) were pump-ed through a sterile hose into a GS 0.22 μm Sterivexfilter unit (Millipore, Maryland, USA), and 1.8 mL oflysis buffer (40 mM EDTA, 50 mM Tris, 0.75 Msucrose) was added to each filter. Filters were frozen at?80 ?C and bacterial genomic DNA extracted as per themethods of Schauer et al. (2000) before being stored at?20 ?C.2.5. PCR amplification of 16S rRNA genesBacterial 16S rRNA genes were amplified withuniversal bacterial primers 27f and 1492r (Lane, 1991)using standard PCR conditions consisting of 5 μL of 10Xbuffer (containing 1.5 mM MgCl2), 2 mM dNTPs, 1.5 UrecombinantTaqDNApolymerase(QIAGEN),10pmolofprimers and approximately 50–200 ng template DNA.DNA amplification reactions were subject to an initialdenaturationat94?Cfor3min,30cyclesof94?Cfor45s,54 ?C for 45 s and 72 ?C for 2 min, and a final extensionstep of 72 ?C for 5 min. For DGGE analysis, PCR wasconducted as per Ferris et al. (1996), with primers 1055f(conserved for domain bacteria; Escherichia coli positions1055–1070)and1392r(universalconservedprimer;E.colipositions 1392–1406) (Ferris et al., 1996) and approxi-mately 100–400 ng template DNA. A 40-bp GC-clampwasattachedtothe5′endofthe1392rprimer(Ferrisetal.,1996;Muyzeretal.,1993).Theseprimersamplifya323bpsection of the 16S rRNA gene of members of the domainbacteria, including the highly variable V9 region. All PCRreactions were performed in either an Eppendorf Master-cycler (Eppendorf, Hamburg, Germany) or a PE AppliedBiosystemsGENEAMPPCRSystem9700(PerkinElmer,Maryland, USA). PCR products were visualized in a 1.5%agarose gel stained with ethidium bromide (0.5 μg mL?1)and stored at ?20 ?C until required.2.6. Clone library construction and analysisAn amplified 16S rRNA gene PCR product (27f/1492r) obtained from a day 10 water sample was clonedusing a TOPO TA cloning? kit, version K2 (Invitrogen,California, USA) as per manufacturer’s instructions. 作文 手机作文网 作文 手机作文网 Atotal of 100 clones were randomly selected and the 16SrRNA gene amplified through PCR as outlined above.Clones were screened by restriction fragment polymor-phism (RFLP) analysis of PCR products using restrictionenzymes HhaI andHAEIII (Promega,Wisconsin,USA).RFLP profiles were visualized on a 3% agarose gel,stained with ethidium bromide (1 μg mL?1) and clonesgrouped into operational taxonomic units (OTUs). Rep-resentative clones of each OTU were selected and grownin Luria–Bertani (LB)-ampicillin (50 μg mL?1) brothovernight.PlasmidDNAwasextractedusingaQIAprep?Spin Miniprep Kit (QIAGEN) as per manufacturer’sinstructions, quantified and directly sequenced.2.7. Denaturing gradient gel electrophoresisDenaturing gradient gel electrophoresis (DGGE) wasperformed using an INGENY phorU?2 (Ingeny Inter-nationalBV) DGGEsystem. PCRproductswererun ona6.5% acrylamide with a 50–70% linear gradient ofurea and formamide and electrophoresed at 60 V, 60 ?Cfor gel 20 h. Denaturing gradient gels were stained with 1XSYBRgoldnucleicacidstain(MolecularProbes,Oregon,USA)for30minanddestainedwith1XTAEfora further30 min. Gels were photographed using a Fluor–S?MultiImager (Bio-Rad, California, USA). Distinct bandswere excised, placed in 100 μL of nuclease-free water(Ambion) and stored at 4 ?C overnight to elute DNA.DNA was re-amplified through PCR as previously des-cribed and DGGE was performed on re-amplified bandstocheckmobilityagainst originalexcisedbands.OriginalPCR products of correct mobility were purified anddirectly sequenced.2.8. Sequencing and phylogenetic analysisPartial and complete 16S rRNA gene sequences ofbacterial clones were obtained using the primers 27f,339f, 732fand 1492r (Lane, 1991).DGGEPCR productswere sequenced with the primer 1055f (Ferris et al.,1996).AutomatedsequencingwasperformedusingeitherAmersham DYEnamic ET-terminator sequencing dye(AmershamBiosciences)orABIBigDyeTerminatorv3.1sequencing dye (Applied Biosystems, CA, US), as permanufacturer’s instructions. Sequences were checkedfor chimera formation with the CHECK_CHIMERA82M.S. Payne et al. / Aquaculture 258 (2006) 80–90 software of the Ribosomal Database Project (Maidak etal., 1996). Sequence data was aligned to the closest re-lative using the BLAST database algorithm (Altschul etal., 1997), and then further analysed with the ARB soft-warepackage (Ludwig etal.,2004).Tree topologieswereevaluated by reconstructing phylogenies using evo-lutionary distance (Phylip Distance Method with Jukesand Cantor model) analysis of aligned near full-lengthsequences (N1000 bp) (Ludwig et al., 1998). Regions ofambiguoussequence(N)wereremovedfromtheanalysis.Aligned partial 16S rRNA sequences (b1000 bp) weresubsequently inserted without changing the overall treetopology using the parsimony tool available within ARB.Bootstrap values were obtained for branching patternsusing the Phylip software package (version 3.65) (Retief,2000) and values ?50% were included for main nodes ofthetree.ThenucleotidesequencedataofallDGGEbandsand clones appear in the GenBank nucleotide databaseunder the accession numbers, DQ334388–DQ334399and DQ334360–DQ334374 respectively.2.9. Statistical analysisFlow cytometry data was analysed using a one-wayanalysis of variance (ANOVA) in SPSS (version 12.1).Homogeneity of variances for the data set was assumed.Standardized residual and predicted values for bacteriacountswereobtainedandgraphedonascatterplottocheckrandom distribution of data. Standardized residual valuesforbacteriacountswerealsographedonahistogramandanormal curve applied to check the normal distribution ofthe data. Data was transformed using a log10+1 algo-rithm. Where significant differences were recorded, datawas further analysed using a Tukey HSD post hoc test.Clone library data was analysed using various 作文 手机作文网 作文 手机作文网 indicesandmodelstoevaluatethevariationofmicrobialdiversitywithin the library (Magurran, 1988). Rarefaction analysis(Hecketal.,1975;Simberloff,1978)wasconductedusingthe analytical approximation algorithm of Hurlbert(1971). Calculations were performed with the freewareprogram aRarefact Win (Holland, 1988). The freewareprogram, EstimateS, version 7.5 (Colwell, 2005) wasused to calculate the following indices: The Shannon–Weaverindex(ShannonandWeaver,1963),calculatedas:HV? ?PSof OTUs (Moyer et al., 1994); Fisher’s alpha log seriesrichness index (Fisher et al., 1943); Chao-1 (Chao, 1984;Chao, 1987) calculated based on the number of species ina sample that are represented by one or two individuals(Foggo et al., 2003) and coverage (C) values, calculatedbytheequation;C=1?(n/N)×100wherenisthenumberi?1pilnepiT where pi is the proportion ofclones belonging to the ith OTU and S is the total numberof unique clones and N is the total number of clonesexamined (Good, 1953).3. Results3.1. Flow cytometry analysisFlow cytometry detected an increase in bacterialnumbers within the water column in the early stages ofthe larval rearing trial. Numbers increased by approxi-mately 52% from day 1–2, 46% from day 4–5 and 24%from day 5–6, before peaking at day 7. Following thispeak bacterial numbers decreased by approximately 52%and remained relatively stable until day 10 (Fig. 1).Statistical analysis of the data, however, demonstratedthere were only significant differences (One-wayANOVA,P=0.00,Pb0.05)inbacterialnumbersbetweendays 1 and 3–10 and day 2 and days 6 and 7.3.2. Denaturing gradient gel electrophoresisThe DGGE bacterial profile from the larval rearingtrial demonstrated a dynamic microbial communitywithin the water column of the rearing tank (Fig. 2). Ashiftinthemicrobialcommunitycompositionfromday1to day 2, corresponding with the post-stocking of thelarval rearing tank was observed. In addition, the bac-terial profile changed markedly between day 4 and day 5though no distinct change in the microbial communitystructure occurred between days 9 and 10, when masslarval mortalities were recorded within the larval rearingtank.The majority of band sequence affiliations obtainedfrom the DGGE analysis resided within the γ-Proteobac-teria,α-ProteobacteriaandFlavobacteriagroups(Table1,Fig. 3). Three bands dominated the bacterial fingerprint onday 1 (Fig. 2), though only two were successfully020040060080010001200140013810DayBacteria (cells/ml)976542Fig. 1. Bacterial numbers present in the water column of the P. ornatuslarval rearing system over the duration of a larval rearing trial asrecorded by flow cytometry analysis.83M.S. Payne et al. / Aquaculture 258 (2006) 80–90 sequenced and closely affiliated with Methylophagaburyatica (band 1) and an uncultured γ-Proteobacterium(band2).AThiomicrospirasp.affiliatedband(band4)wasdominant early in the rearing period, though this band wasnot maintained through the larval run. A Flavobacteriumsp. affiliated band (band 5) dominated the profile on day 4but was subsequently only weakly detected in laterfingerprints (bands 5 and 11). On day 5 a Roseobactersp. (band 8) and a Bacteroidetes (band 9) affiliatedsequenceappeared.Band8remaineddominantthroughoutthe larval rearing trial while band 9 was absent from theprofile on day 8 but reappeared on day 9 and remaineddominant until the cessation of the larval rearing trial.InterestinglytheMethylophagasp.affiliatedband(band1)was again detected throughout the later stages of the trial.3.3. Clone library analysisRestriction fragment length polymorphism (RFLP)analysis of a clone library (100 clones) sampled 作文 手机作文网 作文 手机作文网 followingamassmortalityevent(day10)duringalarvalrearingtrial,identified 59 independent OTU groups. Representatives of6 OTU groups that contained two or more clones weresequenced.ThedominantOTUgroup(OTU1:28%ofthelibrary) was closely affiliated with an uncultured γ-Proteobacterium (DQ189582). OTU 2, representing 7%of the library was closely affiliated with an unculturedBacteroidetes bacterium clone (AY580666). Interestingly,OTUs 3 and 4, which each represented 4% of the librarywere closely affiliated with a Vibrio corallilyticus(AJ440004) and an Acidovorax sp. (Y18617) respectively.Fig. 2. DGGE profile of 16S rRNA gene fragments of water samples taken from the P. ornatus larval rearing system over the duration of a larvalrearing trial. Numbers at the top of the figure (black) represent days. Numbers in the main body of the figure (white) represent the bands that were cutfrom the gel and sequenced. These sequences are present in the phylogenetic tree in Fig. 3. Bands marked with the symbol ‘X’ represent bands forwhich sequence data could not be obtained.Table 1Affiliation of sequences retrieved from DGGE DNA bands (Fig. 2) based upon partial (250–350 bp) 16S rRNA sequenceaBand #Closest relative and database accession numberAlignment (bp)Similarityb(%)Taxonomic description12Methylophaga buryatica (AY128533.1)Uncultured gamma proteobacterium, clone BIrii5(AJ318170.1)Proteobacterium Ellin139 (AF408981.1)Thiomicrospira sp. JB-A2 (AF013974.1)Flavobacterium sp. 10B, isolate 10B (AJ698832.1)Filamentous photosynthetic bacterium Fl90-32(AB046591.1)Sulfitobacter sp. DG885 (AY258079.1)Roseobacter sp. DG942 (AY258088.1)Uncultured Bacteroidetes bacterium clone 1A5(AY274868.1)Rhizobium galegae strain SAFR-030 (AY167831.1)Flavobacterium sp. 3A5 (AF368756.1)Ectothiorhodosinus mongolicum, strain M9(AY298904.1)319/328314/3299795γ-Proteobacteriaγ-Proteobacteria3456116/119303/312320 /329286/28897979799Proteobacteriaγ-ProteobacteriaFlavobacteriaBacteria789140/144295/29 6325/329979998γ-Proteobacteriaα-ProteobacteriaBacteroidetes101112317/320321/337314/32 1999597α-ProteobacteriaFlavobacteriaγ-ProteobacteriaaBands marked X in Fig. 2 did not sequence well and as such, potential identifications were not provided.bSequenceswerealignedto the closestrelative usingBLAST(Altschulet al., 1997). Thesimilarity wascalculatedwithout gapstakeninto account.84M.S. Payne et al. / Aquaculture 258 (2006) 80–90 OTUs 5 and 6 each represented 2% of the library andaffiliated closely with a Vibrio sp. (AF513461) and anuncultured γ-Proteobacterium clone (DQ189582) res-pectively. A further 9 randomly selected clones from thelibrary representing individual OTU groups were se-quenced to provide additional phylogenetic information0.1Clone - CWOTU36 (1%)Vibrio fortis (AJ514916)Vibrio coralliilyticus (AJ440004)Vibrio brasiliensis (AJ316172)Clone - CWOTU3 (4%)Clone - CWOTU43 (1%)Vibrio furnissii (X76336)Clone - CWOTU23 (1%)Vibrio alginolyticus (AY332566)Vibrio parahaemolyticus (AP005082)Vibrio calviensis (AF118021)Clone - CWOTU5 (2%)Clone - CWOTU6 (2%)DGGE - BD4Thiomicrospira frisia (AF013974)Clone - CWOTU1 (29%)Uncultured γ-Proteobacterium (AJ240921)Proteobacterium Ellin139 (AF408981)DGGE - Band 3DGGE - Band 2Uncultured γ-Proteobacterium (AJ318170)Acidovorax sp. (Y18617)Clone - CWOTU4 (4%)Clone - CWOTU54 (1%)Pseudomonas sp. (AY520572)DGGE - Band 12Ectothiorhodosinus mongolicum (AY298904)Clone - CWOTU58 (1%)Uncultured γ-Proteobacterium (AB015514)DGGE - Band 1Methylophaga thalassica (X87339)Clone - CWOTU10 (1%)Candidatus arcobacter (AY035822)DGGE - Band 8Sulfitobacter mediterraneus (Y17387)Roseobacter sp. (AY258088)Clone - CWOTU7 (1%)Clone - CWOTU17 (1%)DGGE - Band 6Ruegeria sp. (AJ295988)DGGE - Band 7Salipiger mucescens (AY527274)DGGE - Band 作文 手机作文网 作文 手机作文网 10Rhizobium galegae (AY167831)DGGE - Band 11DGGE - Band 5Flavobacterium sp. (AF368756)Clone - CWOTU2 (7%)Uncultured Bacteroidetes (AY274855)DGGE - Band 9Rhodothermus marinus (X77140)100100100907510099100100γ-Proteobacteriaβ-Proteobacteriaγ-Proteobacteriaε-P roteobacteriaα-ProteobacteriaBacteroidetesFig. 3. Phylogenetic tree of 16S rRNA gene sequences recovered from the clone library and DGGE analysis of water samples taken from the P.ornatuslarval rearingsystem. The scale barrepresents 0.1 changespernucleotide. GenBankaccessionnumbersare providedfor all clone, DGGE andreference sequences. The proportion of each clone within the clone library is represented in brackets following the clone name. Locations of DGGEbands are found in Fig. 2. Partial clone and DGGE sequences were added to the phylogenetic tree using the parsimony algorithm tool in the ARBsoftware package (Ludwig et al., 2004). Rhodothermus marinus was used as an outgroup for the analysis. Bootstrap values ?50% are represented atthe nodes of branching points.85M.S. Payne et al. / Aquaculture 258 (2006) 80–90 onthemicrobialcommunitywithinthewatercolumnofthelarval rearing system. Three (OTUs 23, 36 and 43) wereaffiliated with Vibrio sp. while two α-proteobacterialSulfitobactersp.relatedsequenceswereidentifiedforOTUgroups 7 and 17. Phylogenetic affiliations for all OTUgroupssequencedareprovidedinTable2withmostfallingwithin the γ-Proteobacteria (Fig. 3).Calculated diversity indices (Shannon–Weaver H′—3.35; Fisher’s alpha—57.66; 500) confirmedhigh bacterial diversity within the library, supporting thelarge Chao-1— number of RFLP patterns detected from the watersample of the larval rearing environment. The percentageof coverage in the library was relatively low at 42%.Library rarefaction analysis also supported a large bac-terial diversity with the curve not approaching a visibleasymptote (figure not shown). All indices confirmed thatanalysisofalargernumberofcloneswouldberequiredtoobtain a better representative picture of bacterial diversitywithin this sample.4. DiscussionLarvalrearingofthe tropicalrocklobsterP.ornatus,ischaracterised by high mortality in early developmentalstages,especiallyaroundmoultperiods.Previousresearchhas highlighted opportunistic bacterial pathogens as onepotential cause of mortality (Bourne et al., 2004). Tobetter understand the microbial challenges faced by phy-llosoma,thelarvalrearingsysteminwhichtheyareraisedhas been broken down into compartments with one com-partmentbeingthewatercolumnandassociatedmicrobialcommunity. This study has investigated the microbialdiversity and quantified the microbial load which phy-llosoma experience in the water column during the earlystages (P1–P2) of a larval rearing trial.Total bacterial load within the water column markedlyincreased in the early stages of the larval rearing trial,peaking at day 7. This time point correlates with thebeginning of the larval moult from phyllosoma stage one(P1) to stage two (P2) and is usually completed in thepopulation within 3 days (Bourne et al., 2004). The de-tected increase in bacterial numbers early in the larvalrearing trial, just prior to the moult may be attributed to aconditioning phase within the water column where bac-teria are exposed to a relatively clean environment andbegin to exploit available nutrient sources. During thisphase,strainsthatareabletoutlilisethedominantnutrientsource/s within the water column multiply, subsequentlybecomingthedominantorganism/s.Earlyconditioningofthe water column was observed in the distinct change ofthe DGGE bacterial profile from day 1 to day 2 and againbetween day 4 and day 5 after which the bacterial com-munity fingerprint appeared more 作文 手机作文网 作文 手机作文网 stable with only smallchanges in community composition.Table 2Affiliation of sequences retrieved from clones based upon partial 16S rRNA sequence comparisonOTU # Composition of library (%) Closest relative and database accession number Alignment (bp) Similaritya(%)Taxonomic description129Uncultured gamma proteobacterium cloneSIMO-2607 (DQ189582.1)Uncultured Bacteroidetes bacterium, clonePI_4q11g (AY580666.1)Vibrio corallilyticus, strain LMG 21349(AJ440004.1)Acidovorax sp. BSB421 (Y18617.1)Vibrio sp. PH1 (AF513461.1)Uncultured gamma proteobacterium cloneSIMO-2607 (DQ189582.1)Sulfitobacter mediterraneus (Y17387.1)Uncultured epsilon proteobacterium, isolateMZ-73.NAT (AJ810558.1)Sulfitobacter sp. KMM 3457 (AY682197.1)Vibrio alginolyticus, strain EcGS021001(AY332566.1)Vibrio pelagius, strain CECT 4202T(AJ293802.1)Vibrio corallilyticus, type strain LMG 20984(AJ440005.1)Pseudomonas sp. WAI-21 (AY520572.1)Uncultured bacterium, clone pIR3BA0(AY354173.1)737/73899γ-Proteobacteria27728/73998Bacteroidetes34748/75299γ-Prot eobacteria456422703/703647/680547/5591009597β-Proteobacteriaγ-Proteobacteriaγ-Proteo bacteria71011931/953706/7259797α-Proteobacteriaε-Proteobacteria172311525/531756/76298 99α-Proteobacteriaγ-Proteobacteria361592/59399γ-Proteobacteria431494/49899γ-Proteob acteria545811778/779495/5499990γ-ProteobacteriaBacteriaaSequences werealigned to the closestrelative using BLAST(Altschulet al., 1997). The similarity was calculated without gapstaken into account.86M.S. Payne et al. / Aquaculture 258 (2006) 80–90 The drop in bacterial numbers from day 7 to day 8, isunlikely due to the exhaustion of the readily availablenutrient source/s through microbial turnover since a 50%daily water exchange occurs. The beginning of the moultpotentially changes the nutrient dynamics within thesystem subsequently lowering availability and reducingbacterial load. However, further work involving the mea-suringofdifferentnutrientlevelswithinthe water columnover the duration of a larval rearing trial is needed toclarify this. Alternatively, previous studies have demon-stratedaprogressiveincreaseinbacterialfoulingoflarvaewithin rearing trials up until the moult shedding (Websteret al., in press). This fouling potentially acts as a seedbacterial population sloughing into the water column.Reduced bacterial load on the external surfaces of thelarvae following a moult may reduce the bacterial loadsloughing into the water column. Although the decreasebetween day 7 and 8 was deemed statistically insignif-icant,anincreaseinthereplicationofwatersamplestestedwould likely change this result by lowering the standarderror.Despitebacterialnumbersdroppingfromday7to8,there were no corresponding changes in the associatedDGGE profile at this point in the trial. The result isindicative of a drop in bacterial numbers, but not with achange in microbial community structure.Sequences retrieved from the current study provide arepresentation of only the free-living bacterial diversitywithin the water column, since bacterial aggregatesassociated with live feeds and larvae were excluded byfiltration during sampling. There was a high microbialdiversity within the water column, as indicated by theShannon–Weaver index (Shannon and Weaver, 1963),Fisher’s alpha log series richness index (Fisher et al.,1943), Chao-1 (Chao, 1984, 1987) and rarefaction ana-lysis (Hurlbert, 1971; Heck et al., 1975; Simberloff,1978) of the clone library. Many of the sequences ret-rieved from this study affiliated closely with marinemicrobial strains involved in nutrient cycling. In par-ticular, sulphur cycling is of some importance, as indi-cated by the retrieval of sequences associated withThiomicrospira sp. (band 4) and Sulfitobacter sp. (OTUs7 and 17; band 7). These results support previousobservations that organisms involved 作文 手机作文网 作文 手机作文网 in sulphur cyclingare relatively abundant in this aquaculture system(Bourne et al., 2004). Thiomicrospira sp. are sulphur-oxidising bacteria that are commonly found in inshoremarine muds and eutrophic habitats (Sieburth, 1979),while Sulfitobacter sp. are sulphite-oxidising bacteriathat are common in marine environments (Gonzalez andMoran, 1997; Pukall et al., 1999). The retrieval ofsequences affiliated with the denitrifying bacteria, Acid-ovorax sp. (OTU4) and the non-methane-utilising,methylotrophic bacteria, M. buryatica (band 1) suggestthe presence of nitrogen cycling and carbon cycling/dimethylsulfide metabolism within the water columnrespectively (Khan et al., 2002; Mergaert et al., 2001;Schloe et al., 2000; Janvier et al., 2003). Largeenvironmental cloning studies have previously inferredwidespread autotrophic accumulation of carbon andsulphur among uncultured marine bacteria through thecommon retrieval of functional genes involved in themetabolic cycling of these nutrients within environmen-tal libraries (Venter et al., 2004).The clone library constructed in this study was from awater sample taken during a mass larval mortality event(day 10). This time point was chosen as analysis mayreflect potential known pathogenic strains being presentin the water column. The dominant OTU group (OTU 1)was affiliated with a number of uncultured γ-Proteobac-teria, all isolated from a range of marine environmentsincluding salt marshes, estuaries (Acinas et al., 2004),salterns of varying salinities (Benlloch et al., 2002), andseawatermesocosms(Schaferetal.,2000).Thisdominantsequence was not affiliated with any bacteria that areknownpathogensofcrustaceans.HoweverfiveadditionalOTUs from the clone library were affiliated with Vibriosp. which have previously been implicated as thecausative agent of mass larval mortality (Webster et al.,in press). In addition, due to the low sample coverage(42%) it is possible that a higher number of sequencesaffiliatedwithVibrio sp.couldhavebeendetectedifmoreclones were sequenced.Rosemarkand Fisher(1988) suggested that Vibrio sp.are part of the normal bacterial flora of lobsters, thoughhighlighted that they also become infectious when thehost is subject to increased levels of stress. Unlike thenatural marine environment, animals reared in the aqua-culture environment are subject to a variety of stressesstemming from high stocking densities and sub-optimalwater conditions. In particular, larvae become stressedduring a moult due to the energy expended during thisprocess, leading to a lower immune function (Handlin-ger et al., 1999). As a result, defense mechanisms inculture animals are commonly adversely affected, and inan environment that is known to harbour numerouspatho... 作文 手机作文网