Mineralogical Magazine, June 2003, Vol. 67(3), pp.563–579

Opal-Aandassociated microbes from Wairakei,NewZealand:the first 300 days

1, 2 3 B. Y. SMITH *, S. J. TURNER AND K. A. RODGERS 1 Department of Geology, University of Auckland, Private Bag 92019, Auckland, NewZealand 2 School of Biological Sciences, University of Auckland, Private Bag 92019, Auckland, NewZealand 3 Research Associate, Australian Museum, Sydney, NSW2000, Australia

ABSTRACT

Allsamples of silica sinter, <2 y oldtaken from the discharge drain of the Wairakei geothermal power stationand the Rainbow Terrace of Orakei Korako, consist of non-crystalline opal-A. This silica phase depositsdirectly upon the concrete drain wall and filamentous templets, extending from this wall, affordedby the microbial community present in the drain, whose nature was determined by a culture- independentstrategy that entailed construction, fingerprinting and sequencing of a 16Sclone library. Thebacterial community is dominated by five major groups of organisms, present in approximately equalproportions, and which account for ~50% of the community. None of the 16S sequences from thesedominant groups yielded a perfectmatch with 16S sequences for named organisms in the internationaldatabases. However one dominant group clusters with Hydrogenophilusthermoluteus , a thermophilicfilamentous bacterium, and two cluster with putatively thermophilic members of the Cyanobacteria andgreen non-sulphur bacteria respectively. Initial opal-A deposits rapidly as agglomerationsof silica nanospheres that, in turn, form chains of coalesced, oblate, microspheres <0.460.2 mmaboutthe barbicel-like filaments, to produce a matof fine woven strands. The majority ofindividual filaments are <8 mmlongand 0.8 mmwidebut may be up to 55 mm long by 1 mm wide. Wherelaminar flow dominates, most strands develop parallel to the drain current but some strands crisscrosswhile others protrude above the mat surface. Where flow is turbulent, strands lack preferred orientationand some adopt a helicalform. In general, following deposition, the values of the scattering broadbandat half (FWHM) andthree quarters (FWTM) ofthe maximum intensity decrease with increasingsample age. The behaviour of the band at one quarter maximum intensity (FWQM) isless consistent,but, in general, the youngest sinters possess the highest FWQM, FWHM andFWTM values thatprove independent of fabric type. Opal-A silica matures following its removal from the parent fluid,especially where the sinter surface is filmed by water. A continualmovement of silica is shown bya secondgeneration of microspheres formed on the silica mat surface, by an increase in size of the initialmicrospheres, and by an increase in maximum intensity of the X-ray scattering broadbands. Similarsilica aging behaviour occurs among young sinters developed upon microbial mats at Orakei Korako.The deposition and aging processes accord with the known behaviour of juvenile opaline silica inboth natural and artificial systems whose pH, temperature and dissolved salt content are similar to Wairakeiand Rainbow terrace: gelling of silica is favoured by the high pH (~8.3) and temperature (~60ºC) oftheWairakei discharge fluid but the high dissolved salt content of the water (Na + = 930 mg/g, Ca2+ = 12 mg/g,Cl =1500 mg/g)and abundant microbial community facilitate rapid and copious flocculationof solid silica within the drain, in contrast to the slower accumulation on the natural sinter terraceat lower temperature (30 –45ºC) fromless saline dilute bicarbonate-chlori dewaters (Na + = 180 mg/g, Ca2+ = 0.2 mg/g, Cl = 400 mg/g,pH = 8.1).

KEY WORDS: sinter,Wairakei, New Zealand,bacterial community, microbes.

*E-mail: [email protected] DOI: 10.1180/0026461036730118

# 2003The Mineralogical Society B.Y. SMITHETAL.

Introduction Theresults presented here constitute a Žrststep towardsobtaining a totalpicture of thesingular and WHEN Žrstdeposited, sinter from near-neutral intimatecharacter of the primary microbe-mineral alkalichloride waters consists of opal-A, a non- connectionin the Ž rstdays of silicadeposition. crystalline,hydrated silica phase (Smith, 1998; Followingdepositionupon the microbi al Herdianita et al., 2000a).Whileall materials that substrate,Herdianita et al. (2000a)demonstrated comein contactwith discharging thermal waters, thatchanges occur in both the physical properties bethey biogenic or abiogenic, can afford a andin the textures of theopal-A silica phase, such potentialsubstrate for the coagulating silica, it thatthe primary depositional textures, including haslong been known that microbes such as the anybacterial templets, are modiŽ ed to varying Cyanobacteriaand Archaea thatthrive in the hot degrees.Ultimately, opal-A transforms with time springenvironments, are particularly effective at intoparacrystalline opal-CT and/ oropal-C, and, mediatingthe deposition of theopaline silica (e.g. subsequently,to microcrystall inequartz. The Weed, 1889a,b).Thesilica  occulatesupon and maturationprocess typically takes ~50,000 y for aboutthe different microbial templets to produce TaupoVolcanic Zone (TVZ) sinters. mouldsof the Ž lamentswith a rangeof distinctive Inorder to appraisefossil textures it isessential texturesthat have become the focus of particular tounderstand the manner and extent of any interestin recent years (e.g. Cady and Farmer, modiŽcations an opal-A mould may undergo. 1996;Farmer, 2000). WhileHerdianita’ s et al. (2000a)modelof sinter Thenature of the microbial communitie s agingis helpful in this respect, all samples of involvedin the silica sinter depositional process opal-Awith an ageof 41ywereplotted by these andthe mechanisms of microbe-mediated silici- authorson a single,year-1 time line, i.e. no Žcationare, as yet, unclear. Much of our lack of indicationwas given as to whether changes understandinghas stemmed from our inability to occurredin the physical character of opal-A culture,and hence identify and study in vitro, the duringits Ž rstyear. In pointof fact,they recorded vastmajority of environm entallyassociat ed arangeof valuesin theproperties of their year-1 microbes.The recent development of culture- samplesthat is greater than any seen among independentmolecular methods provides a solu- samplesa fewyears old. tionto this problem by enabling the nature of Giventhat a supplyof oldersinter, available at microbialcommunities to be determined from Wairakei,was deposited in distinct time brackets bacterialDNA extracteddirectly from environ- overprevious months, the opportunity was also mentalsamples (Amann et al., 1995). 16S takento investigate whether the spread of values ribosomalRNA (rRNA)genes are a useful observedby Herdianita et al. (2000a) re ects targetfor these analyses as they are highly mineralogicalprocesses initiated immediately conserved,being found in all microorganisms, uponor soon after deposition, following the whilealso containing variable sequence regions, entombmentand siliciŽ cation of the bacterial orsignature sequences, that enable taxonomic precursors,i.e. what changes, if any, occur in classiŽcation organisms to the level of species. opalinesilica immediately following deposition. ComparativeDNA sequenceanalysis of 16S Sampleswith numbers preŽ xed AU areheld in genesis nowwidely used for the characterization theDepartm entof Geolog y,Univers ityof ofcomplex microbial communities and for the AucklandPetrology Collection. taxonomicdescription of microorganisms from a broaddiversity of environments including subsur- facescale at theOtake geothermal power station, Sampleoccurrence Japan(Inagaki et al.,1997)and Yellowstone hot Samplesare categorized into 3 groups(Table 1). springcommunities(Blank et al., 1999; Hugenholtz et al.,1998;Inagaki et al., 2001). Group A At theWairakei geothermal power station, New Zealand(Fig. 1), the discharge drain contains both Thickdeposits, typical of streamer microfacies anextens ivemicrob ialcommuni tyand an silica,accumulate at arateof a fewhundred mm/ y unlimitedsupply of fresh opal-A. Consequently, inthe drains used to discharge waste water from theopportunity exists to analyse both the nature theWairakei geothermal power station. These anddiversity of a microbialcommunity, and the drainsconsist of twoadjoining concrete channels initialsilica depositing amongst that community. each1.5 m wide.The build up of silica is such

564 OPAL-A AND ASSOCIATED MICROBES

TVZ TVZ

L. Taupo

0 200 km W

a

i

k

a

t

o

R .

Orakei Korako

Wairakei

Latitude 38.5S L. Taupo lake

geothermal systems delineated by geophysical methods

inferred caldera

0 10 20 30 km Longitude 175.5E

FIG.1. Sketch map, showing the locations of Wairakei and Orakei Korako geothermal Želds within the Taupo Volcanic Zone. Insert: North Island, NewZealand.

thatthe drains become clogged and require feathers.An elongate strand or group of strands cleaningevery six months or so. A central formsa centralshaft or rachisthat can commonly dividerseparates the channels and was originally growup to50 mminlength. From this, numerous intendedto permiteach to be used independently barb-likestrands, commonly 5 –10 mm long, butground subside ncehas caused the hot divaricate.In turn, these subdivide to produce dischargewater level to overtop this divider and evensmaller barbicel-like Ž laments,up to55 mm ithasbecome a primarysite for sinter deposition long,that interlock in a mannerakin to the (Fig.2). Flow rate is currently ~500 l/ swithwater barbicelhooks of feathers. Unlike feathers, coolingfrom 95º C atinputto ~60º C immediately however,the diameters of silica-coated rachi, abovea majordrop structure. barbsand barbicels are similar. Theresulting, silica-coated, streamers posses ForDNA analysis,sinter sample swere anunderlyinghabit that resembles the structure of collectedwhich comprised the growing tips of

565 B.Y. SMITHETAL.

TABLE 1.Representative opal-A scattering band parameters of silica from young sinters of Wairakei and OrakeiKorako, New Zealand.

Group Age Age Location AU no1 Max. Position of FWQMFWHM FWTM (y) (days) intensity max. intensity DAÊ DAÊ DAÊ (c/s) (AÊ )

A 40.12<46 Wairakei drain 52387_6h 93 3.99 2.231.37 0.80 40.13 <49 52387_3d 90 4.00 2.101.29 0.75 40.14 <51 52387_5d 94 3.95 2.171.33 0.77 40.16 <62 52387_16d90 3.98 2.201.39 0.79 40.20 <73 52387_37d92 3.97 2.201.34 0.79 40.29 <106 52387_60d114 4.02 2.131.33 0.71 B 40.4<<<234 Ex Wairakei drain52388b 103 3.93 2.201.33 0.84 40.4 <<<234 522388a 97 3.99 2.231.35 0.79 40.5 <<234 52389b 103 3.95 2.101.32 0.84 40.6 <234 52389a 90 3.94 2.231.32 0.68 C <0.25 Orakei Korako 52391 102 4.00 2.161.40 0.80 <0.5 52396a 104 3.98 2.131.39 0.81 <1 52392a 106 4.00 2.141.30 0.72 ~1 52392b 112 4.02 2.041.25 0.72 >1 and <2 52394 107 3.99 2.031.30 0.76 <2 52393 110 3.97 1.981.30 0.77

1 h=hours;d =days

fresh,we ll-developedstre amermicrofacies haveelapsed since deposition is 42 daysalthough depositedon the central divider. Samples were itseems likely that the tips, at least, of the placedin 15 ml sterile polypropylene tubes, stored collectedsample may be nomorethan a fewdays at4º C overnightand then transferred to –20ºC old.Four days elapsed between removing the storagepending further analysis. A largersample mineralogicalsample from the drain and its initial ofthe same material was taken for mineralogical physicalanalysis. analysis( sampleAU52387).Immediately followingcollection, this sample was placed in a cleancontainer, Ž lledwith the parental discharge water,which, at thetime of sample collection, had atemperatureof62.5º C anda pHof 8.3. Subsequently,subsamples were removed from thewater and air dried at room temperature for differingperiods. These intervals are denoted by thenumerical sufŽ x attachedto the sample numbere.g. _6h = 6hours,_3d = 3days,and representtime elapsed following removal from theparent  uid.A minimumof six hours was neededto air dry the sample sufŽ ciently for Flow direction mineralogicalanalysis and this initial subsample, AU52387_6h,has been taken as thebenchmarkof fresh,unaltered silica sinter for the purposes of 0.5 m this study. AllGroup A sampleswere very young, particularlygiven that the drain had been FIG.2. Silica sinter, displaying streamer fabric, actively cleanedout <6 weeks earlier (R. Smith, pers. depositing upon the central divider of the Wairakei comm.)As such, the maximum time that could discharge drain.

566 OPAL-A AND ASSOCIATED MICROBES

Group B microbesfrom the environment allyprotected Sintersamples from the last clearing of thedrain, OrakeiKorako area all provided insufŽ cient 6weeksearlier, were collected from the adjacent silicafor powder X-ray diffracti on(XRPD) bank.All consisted of massive streamer micro- analysis.This proved to be the case particularly faciesscraped from the walls of the discharge inan attempt to sample very young, living sluice.These samples are identical in all respects streamermicrofacies, AU52404. The bacterial togroupA exceptthey were formed at an earlier mathad formed in a channel,0.15 m wideand dateand have been subject to slight weathering. 0.04m deep,with a watertemperature of 56º C Giventhat the second most recent cleaning of the anda velocityof 0.3 m/ s.All Ž lamentswere drainhad occurred 6 monthspreviously, the alignedwith the  owdirection.Similarly, a very maximumpossible depositional age of any of young,living, pustular orange bubble mat, from theseolder samples, at the time of analysis, was anadjacent area, AU52390, yielded insufŽ cient 234days. Sample AU52389 exhibits a zonal silica.This sample was growing on a mid-slope structure.The zone deposited immediately adja- sinterapron in water 0.01 m deep,with a centto the drain wall (AU52389a, zone 1) has velocityof 0.5m/ sanda temperatureof 42.3ºC. beenassigned an age of <0.6 y, the maximum Adequatesilica was obtained from somewhat timeelapsed from deposition. By deŽ nition the older,but still young, partially siliciŽ ed, Ž lamen- outer(drainward) zones are younger and the next touscyanobacteria-lik eassemblagesrepresenting layerdrainwards (AU52389b, zone 2) was taken bothbubble mat and palisad emicrofacies; ashaving an arbitrary age of <0.5 y. Two samplesAU52391 and AU52396a, respectively. additionalsamples, AU52388a and AU52388b, SampleAU52391camefr omapartially bothexhibit well developed streamer fabric and submergedpustular bubble mat growing on a showno evidenceof havingbeen formed adjacent mid-slopesinter apron in 45º C water,with a tothe wall. They are taken, subjectively, as velocityof 0.3 m/ s.The silica was taken from the havingages of <0.4 y. siliciŽed periphery of the mat where it seemed likelyto have been deposited within the previous threemonths. Palisade sample AU52396a came Group C fromthe lip of a channelat the top of an TheWairakei drain specimens fail to provide a accumulatingsinter apron and was associated totaloverlap in age with those examined by withgreen, living, mid-temperature-size dcyano- Herdianita et al. (2000a).Inorder to conŽ rm and bacteria-likeorganisms. The water temperature morefully explore the nature and extent of trends was 54ºC. observedinthe physica lpropertiesof the Oldersamples AU523924 from the bubble mat Wairakeisilicafollowi ngits deposi tion,a microfaciesprovide the second end-member for numberofa dditionalex amplesof youn g, thepresent study, complementing the very fresh naturallyoccurring silica sinter were taken from Wairakeidrain AU52387. Samples AU52932 thelarge sinter terraces of Orakei Korako, an camefrom a partiallysubmerged, totally siliciŽ ed activegeothermal Ž eldsituated some 18 km bubblemat developed on abroadsinter apron in a northnortheastofWairakeiin t heTa upo watertemperature of 53º C. AU52392ais fromthe VolcanicZone (Fig. 1; Lloyd, 1972). No DNA totallysiliciŽ ed toplamina of the siliciŽ ed bubble analyseswere undertaken of any of these speci- matformed at the air/ waterinterface. The age of mens.Sinter accumulation rates from the dilute thisupper exposed lamina is estimated to be bicarbonate-chloridewater discharging at Orakei >0.5y but<1.0 y. AU52392b consists of smooth Korakoare markedly lower than at Wairakeiand silicalaminaef ormedimmediatelyb elow varyfrom a highof 25mm/ yaboutHochstetter’ s AU52392aand is considered to be ~1.0 y old. Poolto an annual average of ~1mm/ yatGolden SamplesAU52393 –4camefrom dry sinter FleeceTerrace (Lloyd, 1972). Estimates of terracesand consist of smooth silica laminae sampleages have been made, partly on these formedabout voids. Their age is estimated to be ratesof sinter accumulation and, partly on the <2.0 y. positionthe different samples occupied within the Importantly,none of the samples in this group sinteraprons. hadsuffered overprinting from post-depositional Acompleteoverlap of sampleages with those steamheating or weathering,both of which,it was fromWaira keidr ainwa snotpossib leas found,can affect the physical properties of young examplesof ve ryy oung,partlys iliciŽed sinter.

567 B.Y. SMITHETAL.

Methods andPB38 (GKTACCTTGTTACGACTT). These primersamplify the near-full length 16S gene Analysisof themicrobial community in the drain spanningpositions 8 –1509 (E. coli numbering: samplewas undertaken using a culture-indepen- Brosius et al.,1981).Samples were also screened dentstrategy that entailed extraction of bacterial for Archaea usingPCRprimersA16S.1 DNA fromthe sinter samples followed by PCR (CCAGGCCCTACGGGGCGCA) andA16S.2 ampliŽcation, cloning, restriction fragment length (GTGTGCAAGGAGCAGGGAC).NoPCR polymorphism(RFLP) Žngerprintingand sequen- productswere detected using these primers and cingof a16Sclone rRNA genelibrary. A detailed allfurther analysis was centred on PCR products reviewof this and other 16S-based methods is obtainedfrom the bacterial primers. The PCR presentedin Amann et al. (1995).Unless reactionswereperformed in 50 ml volumes otherwisestated in thesummary below, molecular containing20 m M Tris-HClpH 8.6, 50 m M KCl, methodswere those described in Sambrook and 2.0 mM MgCl2, 100 mMeachdNTP, 1.0units Russell(2001), with DNA extractionperformed PlatinumTaq TM (InvitrogenLife Technologies, usinga modiŽcation of a methoddescribed by CA), 0.2 mM ofeach primer and 2.0 ml of DNA Miller et al. (1999).Full experimental details may template.AmpliŽ cation was carried out using a beobtained from one of the authors (SJT). programof 94ºC for3 minfollowed by 25 cycles of94º C, 45s; 55º C, 45s;and 72º C, 90s. DuplicatePCR reactionswere carried out on DNAisolation eachsample. The duplicate PCR productswere Approximately0.5 g ofthe sinter sample was pooledand then puriŽ ed using a HighPure PCR weighedinto a sterile2 mlscrew-cap microtube PuriŽcation Kit (Roche Diagnostics, Mannheim, containing0.5 g eachof 0.1 mm and3.0 mm Germany)accord ingto the manufa cturer’s silica-zirconiumbeads. Equal volumes (300 ml) of instructions.Negative (no DNA) andpositive phosphatebuffer (100 m M NaH2PO4 pH 8.0), (seeded)controls were included to check for DNA lysisbuffer (100 m M NaCl, 500 mM Tris pH 8.0, contaminationof reagents and the efŽ ciency of 10%SDS) andchloroform-isoamyl alcohol (24:1) PCR ampliŽcation respectively. PCR ampliŽca- wereadded and the tubes machine shaken tionproducts were assayed for quality and vigorously(45 m/ s)for 40 s. The tubes were quantityby polyacrylamide gel electrophoresis thencentrifuged at 13,100g for5 minand 650 ml (PAGE) againsta Low DNAMassLadder ofsuperna tantwere recover edfrom which (InvitrogenLife Technologies). proteinsand SDS wereprecipitated by adding of 360 l of 7 M ammoniumacetate and placing the m Clonel ibraryconstruct ionand sc reening tubesonic efor5min.Thetubeswer e centrifugedat 13,100 g for5 minand the upper PuriŽed PCR productswere ligated into a T-tailed aqueousphase recovered and DNA precipitated PCR cloningvector (pGEM-T; Promega Madison byadding 0.54 volumes of isopropanol ,and WI)essentiallyas describedin the manufacturer’ s centrifugingagain at 13,100 g for5 min.The protocols.Ligation reactions were prepared in supernatantwas discarded and the pellet washed 15 mlvolumesand incubated overnight at 4º C. with1 mlof 70%ethanol, followed by centrifuga- Theresulting ligation products were transfected tionat 13,100 g fora further5 min.Following into E. coli DH5a competentcells in accordance removalof thesupernatant, the pelletwas allowed withthe manufa cturer’sprotocolwith the toair dry for 30 min then resuspended in 100 ml exceptionthat Luria broth was used instead of ofwater. To improve recovery of DNA the SOC medium.Transformants containing inserts extractswere supplemented with 2 ml of linear wereidentiŽ ed byblue-whiteselection on L-agar polyacrylamidecarrier (Molecu larResearc h plates.A clonelibrary of 96 randomly selected CenterInc., Cincinna tiOH) priorto DNA whitecolonies was prepared for further analysis. precipitation. Individualclones were subcultured in 150 ml of L-brothin a microtitre-trayformat and incubated at37º C overnight.Crude cell lysates were PCRamplif |cationof 16Sgenes preparedfor PCR analysisby transferring 50 ml Bacterial16S rRNA genesequences were PCR ofthe broth culture to a newtube containing ampliŽed from the sample DNA usinguniversal 100 mlofsterile H 2Oandincubating at 94º C for primersPB36 (AGRGTTTGATCMTGGCTCAG) 20min. Lysates were centrifuged at 3000 g for

568 OPAL-A AND ASSOCIATED MICROBES

5minto sediment cell debris and were stored at (http://www.ncbi.nlm.nih.gov)using the program –20ºC priorto analysis. The remaining portions BLAST (Altschul et al., 1990). ofthe cultures were supplemented with 15% glyceroland stored at –80ºC. Insertswere Mineralogy recoveredfrom cell lysates by PCR ampliŽcation usingvector-speciŽcprimerspGEM-F Ofthe different physical properties of silica sinter (GGCGGTCGCGGGAATTCGATT) andpGEM- phasesthat Herdianita et al. (2000a)foundto vary R(GCCGCGAATTCACTAGTGATT). ThePCR withtime, the most sensitive response was in the reactionswereperformed in 20 ml volumes sizeand shape of the X-ray scattering broadband containing20 m M Tris-HClpH 8.6, 50 m M KCl, ordiffraction line centred at ~4 A Ê . The most 2.0 mM MgCl2, 100 mM eachdNTP, 1.0units convenientproperty of the opal-A scattering band TM PlatinumTaq , 0.2 mM eachof primerspGEM-F usedby Herdianita et al. (2000a,b) to compare andpGEM-R and1.5 mlofcrude cell lysate. individualspecimens was the full width at half AmpliŽcation was carried out using the same maximum(FWHM). Theauthors equated this programas described above for primers PB36 and measurementwith the degree of lattice disorder, PB38.The resulting PCR productswere screened anapproachanalogous to that used to evaluate the forsimilarity by RFLP Žngerprintingusing the crystallinityof someclays. The same strategy has restrictionendonu clease HaeIII.Restriction beenadopted here to appraise differences in digestswere pe rformedin25 ml volumes opal-Aat various stages of its aging process and comprising20 mlofPCR product,2.5 ml of toenabledirect comparisons to be made with the 106 React2 bufferand 5 U HaeIII and were Herdianita et al. (2000a)maturationmodel. incubatedovernight at 37º C. Digestionproducts However,in the present study it was also found wereresolve dbyPAGE through6% non- usefulto include measurements of full width at denaturinggels.Fingerpri ntproŽ les for the quartermaximum (FWQM) andfull width at clonelibrary were compared for similarity and threequarter maximum (FWTM), alongwith the eachdifferent pattern assigned an Operational positionand intensity of the scattering maximum TaxonomicUnit(OTU) referencenumber . abovebackground, in order to better describe the Representativeclones were randomly selected sizeand shape of thedifferent scattering bands for fromthe Ž vedominant OTUs forDNA sequen- comparativepurposes (Fig. 3). cing.Insert DNA wasprepared for sequencing by Allsamples were scanned by a PhilipsX-ray PCR ampliŽcation from cell lysates as described DiffractometerŽ ttedwith a graphitemonochro- above,with the exception that duplicate reactions matorwith acquisition controlled by Diffraction werecarried out in 50 mlvolumesfor each clone TechnologyVisXRD software,using the same tobe sequenced. The duplicate PCR products rigorouslyconstrained instrumental settings and werepooled and puriŽ ed using a HighPure PCR experimentalconditions detailed by Herdianita et PuriŽcation Kit according to the manufacturer’ s al. (2000b)topermit direct comparison between instructions.The DNA sequencingwas performed thepresent results and those of Herdianita et al. usingdye-labelled terminators on an ABI 3100 (2000a).Operatingconditions were 40 kV at Ê sequencerŽttedwith80cmcapillaries. 20mA, usingCu- Ka radiation (la1 = 1.5405 A). Sequencingwas carried out using both forward Samplesof untreated,dry, hand ground, <106 mm andreverse pGEM primersand an internalprimer powders,were mounted in a 10 62061.5 mm 16S.5(GCTCGTTGCGGGACTTAACC) which aluminiumholder, and scanned at 0.6º 2 y/min enabledthe sequence of the entire insert to be from 10 –40º2y,witha stepsize of 0.01º . An obtained. internalsilicon standard was used from time to timeto check the alignment of thediffractometer. Measurementsof FWQM, FWHM andFWTM Sequenceeditingand phy logeneticanalysis havea precisionof ± 0.008 Dd AÊ or ±0.07 D2yº Thesequence data for each clone were compiled, (cf.Herdianita, 2000 b).Themeasurement repeat- checkedfor ambiguities and edited into a full abilityof maximum intensity was no better than lengthconsensussequenceusingthe ±2 c/s. Autoassemblercomputerprogram(ABI). Alongwith the variation in physicalproperties, Preliminaryinformatio nonthe phylogenet ic progressivechanges are known to occur in the placementof sequences was obtained by compar- morphologyof the diffe rentultras tructural isonwith sequences in the Genbank Database texturalcomponents of the sinter. Each of the

569 B.Y. SMITHETAL.

Maximum intensity

160 ) s / c (

120 y t i

s FWTM n e t 80 Smoothed trace n FWHM i

w

a FWQM 40 R Base line

15 20 25 30 35 4 0 °2q

FIG.3. Typical opal-A X-ray scattering broadband showing measurement of full width at half maximum (FWHM), quarter maximum (FWQM)and three-quarter maximum (FWTM)of smoothed trace, along with maximum scattering intensity. All measurements are made relative to abase line constructed as atangent to the smoothed scattering trace.

sampleswas examined by a highresolution DiscreteproŽ les were deŽ ned as Operational scanningelectron microscope (SEM) tosee if TaxonomicUnits (OTUs). The number of OTUs anyconsistent changes were apparent in the Ž rst providesa measureof the diversity of 16S yearof the existence of opal-A. In addition, a sequencesin the clone library and, putatively, cryostagewas used to examine samples AU52404 thebacterial diversity in the original sample. andAU52390 to ascerta inthe relatio nship Similarly,the number of cloneswithin each OTU betweenthe living microbial ultrastructure and providesan indication of the abundance of a initialopal-A  occulant. groupof similar sequences in the library and, putatively,the abundance of aparticularorganism inthe sample. Phylogenetic comparison of clone Results sequenceswith 16S sequence databases provides Microbialcommunit yanalysis anindicationof thetype of organisms represented Atotalof 72 clones from the bacterial 16S library bythe sequenced clones. werescreened by RFLP Žngerprintanalysis The70 RFLP proŽles from the clone library yielding70 unambiguous Ž ngerprintproŽ les. clusterinto 35 distinctOTUs. Twenty-nineOTUs

FIG.4. Representative Restriction Fragment Length Polymorphism (RFLP)proŽ les obtained from the 16S clone library prepared from the discharge drain sinter sample. For the purposes of comparison, each proŽle type was deŽned as an Operational Taxonomic Unit (OTU). The numbers at the top of the Žgure indicate the proŽles of the Žve dominant OTUs(OTU1 –OTU5)that were represented by multiple clones in the library, and for which DNA sequence data were subsequently obtained. Mdenotes the molecular weight size standard (50-bp ladder, Gibco- BRL)and numbers along the side of the Žgure indicate size in base-pairs.

570 OPAL-A AND ASSOCIATED MICROBES

wererepresented by two or less clones in the aligned,albeit poorly, within a candidatedivision library.The remaining six OTUs accountedfor 38 ofenvironmental clone sequences (OP10) of the (54%)of theRFLP proŽles (see Fig. 4). The DNA lowG+C Grampositive bacteria. Many of the16S sequenceinformation was obtained from clones sequenceswithin this divisio nderivefrom representingŽ veof the six dominant OTUs hydrothermalenvironm ents.Given the poor (designatedOTU1– OTU5). The results of the alignmentand signiŽ cant difference in sequence, RFLP analysisand DNA sequencealignments OTU4couldrepresen tanewand as yet obtainedfrom a BLAST searchof the Genbank undescribedbacterial species. OTU 5aligned sequencedatabase are summarized in Table 2. No with the Cellvibrio groupof t heg amma exactsequence matches were detected in the Proteobacteria . Cellvibrio aredescrib edas databasesearches although the sequence from cellulolyticbacteria involved in the degradation OTU2 wasnearly identical (99% similarity) to the ofnaturalcellulosic Ž bres(Blackall et al., 1985). 16Ssequence of an organism described as a Thereis no record of there being thermophilic thermophilicand Ž lamentousmember of the membersof this group. greennon-sulphur bacteria (GenBank accession WhileARDRA analysisprovides an indication numberAB067647). ofthecomposition of themicrobial community of OTUs 1and3 alignedwith groups of as-yet- thesinter sample, the physical arrangement or unculturedorganisms within the Hydrogenophilus architectureof thecommunity remains unknown. groupingof the beta Proteobacteria , and the Itis hence not possible to specify which of the Cyanobacteria ,respectively.BothOTU organismsafford a suitablesubstrate for the sequencesclustered with 16S sequences obtained occulatingsilica or, even, which are sessile and fromthermophilicenvironments.Howeve r whichare not. neitherOTUs werevery closely aligned with organismsthat exist in culture therefore limiting Mineralogy furtherinferences about the likely morphology of theseorganisms. Allsamples showed the characteristi cX-ray Thesequence representing OTU 4wasnot scatteringbroadband of opal-A – thedominant closelyrelated to any existing sequence but phasepresent (cf. Smith, 1998; Herdianita et al.,

TABLE 2.Phylogenetic afŽ liation of 16S sequences from dominant clone types (OTUs) with sequences in the GenBankdatabases.

RFLP typeLibrary GenBank Closestrelative or Identityd Phylogenetic (OTU)a %b AccessionNo. sequencec grouping

1 8.6 AY222297Uncultured Hydrogenophilus sp.1425/ 1437(99%) b Proteobacteria AF402975.1e 2 10.0 AY222298Anaerobic Ž lamentousbacteria 1397/1400(99%) Chloro exi (green AB067647 non-sulphurbacteria) 3 7.1 AY222299Uncultured cyanobacterium 1198/ 1258(95%) Cyanobacteria AF4456777 4 11.4 AY222300Uncultured bacterium 1147/1244(92%) Candidate division OP10 AJ271048.1 5 11.4 AY222301 Cellvibrio sp.AJ289164.1 1397/1426(97%) g Proteobacteria 6 5.6 notsequenced 7, 8, 9 2.8 each notsequenced 10– 35 1.4 each notsequenced a Operationaltaxonomic unit b Percentage ofclones in librarywith the designated OTU c Determined fromthe highest scoring match ina BLASTsearch ofGenBank database d Numberof identicalbases between thequery and GenBank sequence e Obtainedfrom a hotspring at KuirauPark, Rotorua, New Zealand

571 B.Y. SMITHETAL.

a Flow direction b

2 um 2 um

FIG.5. SEMcryostage images of microspheres of juvenile opal-A clustering upon Žlamentous bacteria, Rainbow Terrace, Orakei Korako: ( a)streamer fabric, AU52404; ( b)bubble mat, AU52390.

2000a,b).Opal-Ais the Ž rstsilica phase to andFWTM valueswhen compared with all other depositand, in all cases examined, the youngest samplesimplying that changes occur in the silicahas  occulatedupon a bacterialsubstrate opalinesilic anoton lyimmedi atelyitis (e.g. Fig. 5a,b). removedfrom the parenta lwaterbut also Ingeneral, values of FWHM andFWTM followinginitial depositio n(cf.Iler, 1979). decreasewith increasing sample age (Table 1). Overall,the halfwidth values of the Group A FWHM decreasesfrom 1.37 A Ê at0.12 y oldto andB samplesand the younger Orakei Korako 1.30 AÊ ata ge~2 y ,withFW TMvalues silicas,AU52391 and AU52396a, lie towards the decreasingfrom 2.23 to ~2.0 A Ê overthe same upperend of the range of the youngest samples timeinterval. The behaviour of FWQM valuesis citedby Herdianita et al. (2000a).Importantly, lessconsistent. Samples AU52387_3d and _5d, FWQM, FWHM andFWTM valuesare indepen- showa largerproportional decrease in FWHM dentof fabric type (streamer, palisade, bubble

200

160 d ) s / c (

y e t

i 120 s

n a e t

n c i 80 w a R 40 b

15 20 25 30 35 40 °2q FIG.6. Representative, smoothed, X-ray powder scattering bands for young sinters from the Wairakei Power Station discharge drain ( a –c)and from Rainbow Terrace, Orakei Korako ( d –e): (a)fresh sinter, air-dried for 6h, <46 days old (0.12 y), AU52387a, maximum intensity 93 c/s; ( b)sinter air-dried 16 days, 62 days old (0.16 y), AU52387d, maximum intensity 90 c/s; ( c)sinter from the bank of the Wairakei drain, <<234 days old (<<0.5 y), AU52389b, maximum intensity 103 c/s; ( d)palisade microfacies (<0.5 y), AU52396a, maximum intensity 104 c/s; ( e) bubble mat microfacies (>1 and >2 y), AU52394, maximum intensity 107 c/s.

572 OPAL-A AND ASSOCIATED MICROBES

mat),being similar among all samples of moreor frombarbicel- likeŽ lamentsconsistin gofa lesssimilar age, within experimental error. coatingof smooth silica presumably deposited Ingeneral, the values of maximum intensity aboutthe living microbes. These strands are increasewith increasing age (Fig. 6, Table 1). commonlyaligned parallel with the  owdirection Themaximum intensities of Group A samples (Fig. 7a),butsome divaricate and criss-cross to <0.2y oldshow values <94 c/ sand,apart from forma weftto the main strands warp. In places, AU52389a,are considerably lower than all other occasionalstrand ends protrude above the surface samples.Samples from all three groups within the ofthe mat. Typically, individual barbicels are agerange of >0.2 and <0.5 y havevalues of <8.0 mmlongbut some grow as large as 55 mm, 100±4 c/s.The majority of samples with an age andwhiletheirdiameteriscommonly >0.5y showthe highest maximum intensity 0.3 –0.7 mmitmay be up to 1.0 mm. At high valuesranging from 106 to 112 c/ s. magniŽcation, >50,000 6,theinitial completed strandscan be seen to be formed by agglomera- tionof cluste rsof nanosp hericalparticles Silicamorpholo gy <300nm indiameter. Group A Withdeposition of further silica, the barbicel- TheŽ rst-depositedsilica of theWairakei drain strandstransform to become a stringof juxta- formsa densely-packedmat of opal-A woven posed,coalesced, oblate opal-A microspheres,

FIG.7. SEMimages of sinter, Wairakei discharge drain, ( a,b,d)Group Afresh sinter, AU52387, air-dried for 3days i.e. <49 days old; ( c)Group Bsinter, AU52388, ex-drain <0.4 yold: ( a)densely-packed mat of silica-coated, barbicel-like strands, generally aligned with the ow direction. Some strands crisscross forming aweft to the warp of the main strands; ( b)closely-packed barbicel-like silica strands transforming into strands of oblate microspheres 0.260.4 mm(arrow). Clusters of oblate, opal-A microspheres, 0.2 60.4 mm, have enveloped some smooth silica strands; (c)tangled chains of oblate, coalesced spheres forming aclosely-packed mat of barbicel-like strands with clusters of small (~0.4 mmdiameter) late-stage (second generation) microspheres distributed on the surface; (d)twisted, clockwise, helical, silica strands such as occur in both Group Aand Group Bsamples.

573 B.Y. SMITHETAL. commonly0.2 60.4 mm (Fig. 7b).Onlyin a few givingthe sinter its coherency. An undulating and ofthese is the underly ingsubstru ctureof gradationalcontact occurs between this wall- agglomeratednanospheres obvious. Additional attachedlayer and the next silica layer out from silicadeposition can impart an irregular and thewall (AU52389b, Zone 2) which varies in lumpyappeara nceto the strands (Fig. 7 c). thicknessfrom 2 –10mm. Silicain this second Elsewhere,a secondgeneration of occasional, horizonalso occurs in strands but these lack well-formed,~0.4 mmdiameter,opal-A micro- preferredorientation. There is no cementation spheresare scattered across the mat surface. betweenthem. They are bound together by Voidswitha maximumsizeof5x 2 mm, numerousintertwinings and overlaps. Beyond commonlyoccur where strands divaricate at Zone2, the sinter again takes on the typical anglesof up to 45º from the principal direction streamerfabric appearance aligned with the  ow ofstrand alignment – andthe direction of water direction(Zone 3). Individual main central strands  ow. areup to 50 mm long. Euhedralhalite crystals, <1.5 mm along the Thesedrain samples indicate the profound cubeedge, litter the surface of the SEM mount effectthat current  owcan exercise on the habit andpresumably have crystallized from residual ofstreamer microfacies sinter (Fig. 8). Initially, parentalkali chloride  uidupon air drying of the alongthe newly-clea nedrelativel y-at drain sample.Diffraction lines of halite occur in XRPD surface,the hydrodynamics of the drain produce scansof all Group A butnot in any Group B athinlaminar sublayer in the moving  uid(cf. samples,where any traces of theparent  uidhave Blatt et al.,1980).Bacteria populating this presumablybeen washed out by rain. sublayerproduce strands (Ž laments) aligned and orientedmoreor less with the main  ow Group B direction.Silica deposited upon them adopts the Theopal-A of these samples displays a similar habitof the microbial template (Zone 1). Any habitto thatof the Group A streamermicrofacies. irregularitiesin Zone 1 thatproject through the Opal-Aspheres again form chains of oblate, laminar owsublayer can shed eddies into the coalescedmicrospheres, but the average sphere main ow.Differential silica deposition in and sizein Group B samplesis upto1.8 mm diameter aroundthese eddies will yield the undulations andtheir sphericity is more pronounced than in seenin the gradational contact between Zones 1 GroupA suchthat the Group B strandsoften and2 andproduce a surfacethat is hydrodyna- resemblea stringof beads. In places the outlines micallyrough and, in turn,generating a turbulent ofindividual spheres that form the silica strands sublayerwithin the  owpattern. Few microbial havebecome obscured by a latesilica overgrowth strandsthat grow in this sublayer would be (Fig. 7c). expectedto be oriented with respect to the main Asin Group A, asecondgeneration of well- owdirection, nor would any silica moulds that formedopal-A spheres, up to 0.5 mm across, nestlein hollows of theprimary silica mat. Their occurrenceindicates that silica continues to Discharge flow depositafter the sinter is removed from the drain,presumably from inter-pore  uid.It is streamer 3 fabric amongthe oldest Group A andthe Group B samplesthat a highermaximum intensity of the 2 2-10 opal-AX-ray scattering band occurs and this mm increaseis takenas denotingthe initial maturation 12 mm ofthe juvenile silica that occurs within the Ž rst 1 300days following deposition. wall

Wallrock FIG.8. Schematic interpretation of the effect of SampleAU52389a was deposited directly on turbulence on microbial growth habit and silica deposi- thewall of the Wairakei drain. In the horizon tion in Wairakei discharge drain. Zone 1: initial sinter immediatelyadjacent to the drain wall (Zone 1), substrate. Zone 2: secondary substrate that both results i.e.the oldest, closely-packed silica strands are from and causes local turbulence. Zone 3: silica strands alignedboth with one another and the  ow piercing turbulent sublayer, largely aligned with and direction;the close-packing of the silica strands exing in main ow.

574 OPAL-A AND ASSOCIATED MICROBES

mightformaboutthem(Zone2).The (Fig.9). Elsewhere ,individualmicrosph eres pronounced,undulating contact between Zones 2 occurin botryoidal clusters. The proportion of and3 resemblesthat produced during deposition well-formedopal-A sphere sincreaseswith inturbulent uidsuch as iscommonly developed increasingage. downstreamof ripples(cf. Blatt et al.,1980).It is inthis zone, presumably, that helical strands of silicacould grow. However, a numberof bacteria Discussion (e.g.Spiroc hetes)a rekn ownto exhibi ta Molecularcommunityanalysis provides the spirillum-likemorphology akin to that illustrated opportunityto gain new insights into the nature in Fig. 7d irrespectiveof  owenvironment (Holt ofmicrobial communiti esthat are typically et al.,1993).Insofar, as the velocity gradient in recalcitranttoanalysisbyculture-based turbulent owsis steeperclose to the channel wall methods.The 16S sequences detected in this thanin the channel centre, turbulence has least studydid not exactly match those of named effecton the owaway from the rough wall, and organisms,and hence the speciŽ c identiŽcation henceon the structure of any potential microbial andfurther characterization of organisms present substrategrowing in this zone, that, in turn,could inthe sample was not possible, nor can speciŽ c inuence the habit of the silica depositing here. DNAproŽles be associ atedwith potent ial Asindivi dualmicrobi alstrands grow they microbialsilica substrates. Similar difŽ culties eventuallypenetrate the turbulent sublayer and wereencountered by Inagaki et al. (2001) in emergeinto the main (laminar) current. This providinga comprehensiveanalysi softhe occursin theWairakei drain, once strands exceed complexbacterial community of Steep Cone hot 22mm. Itis at this point that the strands of spring,Yellowstone. However the phylogenetic streamerfabric again evolve parallel to the placementof the Wairakei sequences provides a principal owdirection (Zone 3). strongba sisfor inferr ingthe divers ityof organismspresentand in some cases their Group C physiologicalproperties. Opal-Aspheres in the >1 and <2 y oldsinter of Preliminaryattempts to PCR amplify16S OrakeiKorako Group C samplespossess the sequencesusing Archaea-speciŽc primerswere habitstypical of those found in innumerable unsuccessful.This does not necessarily indicate youngsinters where silica has deposited upon a theabsence of Archaea inthe sample as it is microbialsubstrate (e.g. Herdianita et al., 2000a; possiblethat the primers did not match the Rodgers,2000). Well-formed, opal-A spheres, organismspresen t.Use of more extens ive <1 mmindiameter, are commonly observed primersets and in situ hybridizationwith coalescedaboutmicrobial Ž lamenttemplets domain-speciŽc probescould have been used to a b

2 um 2 um

FIG.9. SEMimages of young sinter, <2 yold, exhibiting typical juvenile habits, Orakei Korako: ( a)coalescence of opal-A microspheres <1.0 mmin diameter, to form amould about siliciŽed microbial trichomes (arrowed), AU52393; (b)AU52392a, newly-formed opal-A microspheres <0.5 mmin diameter forming botryoidal clusters on Žlaments. Some Žlaments display coalesced, oblate opal-A spheres (arrowed) similar in habit to those in Fig. 7.

575 B.Y. SMITHETAL. validatethe results although such work was moreinformationto be o btainedon th e beyondthe limited resources of thecurrent study. morphologyof theseorganisms and their physical TheRFLP analysisof the Wairakei sinter arrangementwithin the sinter fabric. An indica- samplesuggests that the bacterial community is tionof the types of organisms present in the moderatelydiverse, but is dominated by Ž ve samplesalso provides valuable information on the majorgroups of organisms, present in approxi- possibleculture conditions needed to isolate matelyequal propor tions,that accoun tfor representativeorganisms from these environ- approximatelyhalf of the community. On the ments. basisof phylogeneticinference, it isprobablethat Thedetection of 16S sequences in this study atleast one of these domina ntgroups of thatare most closely related to clone sequences organisms,represented by OTU 2,areŽ lamentous derivedfrom thermophilic environments else- bacteria,and that three others (OTUs 1,3, 4) are wheresupports the notion that these environments thermophilic.The results of the molecul ar comprisea distinctivebacterial community that is analysisare therefore consistent with the thermo- poorlycharacter ized.(Inagaki et al., 2001). philicenvironment from which the sample was Importantly,the dominance of sequences that collectedand ultrastructural observations by SEM representclades of as-yet uncultured organisms ofthe sinter samples which indicate a signiŽcant furtherreinforces the observation that culture- abundanceof Ž lamentoustemplets. basedstudiesdonot provid eanaccura te However,the types of organisms detected in representationof the micro ora in these environ- thisstudy clearly differ from the results of the ments. similarstudies by Inagaki et al. (1998,2001) DifŽculties exist in endeavouring to ascribe whichindicated an abundance of Thermus and speciŽc silicasinter textures to biogenic or Hydrogenobacter sppin sinter-associated assem- abiogenicprocesses (cf. Inagaki et al., 2001). blagesfrom Yellowstone National Park and a However,the deposition and development of geothermalpower plant in Japan. None of the opal-Aand the ensuing changes observed in the clonesequences obtained in our study clustered crystalchemistry and texture following deposition withinthese groups. While it ispossible that this andsubsequent removal from its parent  uid,are isdueto subtlemethodological differences in the consistentwith what is knownof the behaviour of extractionand analysis of microbial DNA from silicain both natural and abiogenic laboratory samples,it is more likely that these differences systemsof comparable pH, temperature and reect the effect of different temperatures on the dissolvedsaltcontent to the Wairakei and microbialcommunity structure. The temperature OrakeiKorako parent  uids.At bothlocalities recordedat our sample site was 62.5º C whilethe severaldistinct stages of silica deposition are sintersamples examined by Inagaki et al. (2001) involved:(1) silica nanospheres  occulatefrom werecollected from sites with higher tempera- thesol phase and agglomerate as microspheres; turesranging from 75 to 85º C. Itis also possible (2)the microspheres settle on suitable substrates, thatthere are biogeographic differences in the includingbacteria, and grow in number and size; sintercommunities as the sequence for OTU1 was (3)the original bacteria become siliciŽ ed; (4) mostclosely related to a 16Ssequence that was secondarymicrosph eresform and settle on obtainedfrom a hydrothermalpool in Kuirau substrates;(5) silica overgrowths obliterate the Park,Ro torua,NewZealand(Sunnaand outlinesof the juvenile textures. Bergquist,2001) suggesting that this is a locally Forsilica to  occulatefrom the Wairakei prevalentorganism. dischargewater under the prevailing pH and Somecaution must be applied in the inter- temperatureconditions, the presence of other pretationbacterial abundance based on OTU data agentssuch as dissolved Ca 2+ and Na+ or asthe process esof DNA extraction,PCR microbeswould appear to be essential. Above a ampliŽcation and clone library construction may pHof6, puresilica sols are relatively stable. All allbiasth eratioof or ganismsd etected. silicaparticles ,includingoligomers ,beara Neverthelessthe information obtained in this negativecharge and mutually repel one another. studyprovides the basis for preparati onof Withincreasing pH, there is a decreasein the taxon-speciŽc probesthat can be use din numberof interparticle collisions, and hence in combinationwith direct microscopic methods to opportunitiesfor adhesion and growth (Iler, conŽrm the abundance of organisms detected by 1979).Coagulation, i.e. growth of such particles thisstudy. Such future studies will also enable thatmay lead to  occulation,occurs readily

576 OPAL-A AND ASSOCIATED MICROBES

whereparticle-particle interaction is afforded by toa certainsize, there is very little change in residualbonds or some other form of particle- energycontent with surface area and there is an particlebridging. This can occur in the presence optimumupper limit on thesize to which particles ofcoagulants such as salt cations or organic maygrow, at least in the initial stages of molecules.Iler (1979, p. 376) states that, ``in all deposition(Iler, 1979). Further, when the growth cases, occulationis due to interparticle bonding processcontinues, following adhesion, larger throughthe cations ’’.Heillustrates this with particlesgrow at the expense of smaller and the speciŽc referenceto the role that Na + can play latterare consequen tlyeliminate d.The two (Fig.4.17a, p. 377), arguing that coagulation can processes,acting in tandem, produce a limited occuras soonas sufŽcient ion-exchanged Na + has sizerange in the microsphere end product. beenabsorbed on to the surface of each silica Insofaras depositionoccurs preferentially upon particle.Given their similar silica levels, Wairakei thosesurfaces having minimum radius, such as 480 mg/gandOrakei Korako 400 mg/g, it is wheretwo particles are in contact (Iler, 1979), a presumablythe elevated salinity level in the continuumof texturesbecomes possible as isseen Wairakeidischarge  uids(Na + 930 mg/g, Ca2+ intheWairakei samples: individual oblate micro- 12 mg/g, Cl 1500 mg/g,pH 8.3), along with its spheres ? juvenilestrands ? beadstrings ? highmicrobial content, that facilitates  occula- irregularlumps. Growth continues readily wher- tionand rapid build up ofthe opaline silica in the everan aqueous Ž lmis present on the sinter draindespite the high pH (~8.3) and water surfacesuch as may be supplied by pore water, temperature(~60º C) thatfavour gelling. In eitherthermal or rainwater,in thecase of Group B contrast,at Orakei Korako where the parent samples.This Ž lmallows activity gradients to uidis dilute (Na + 180 mg/g, Ca2+ 0.2 mg/g, Cl becomeestablished and maintained that can 400 mg/g,pH 8.1)and discharge levels markedly promotethe movement of silica (Landmesser, lower,sinter grows more slowly, particularly 1995).Importantly, unlike other metal oxides, distalfrom the vent on the sinter aprons. solidsilica remains non-crystalline throughout Subtlevariations in dissolved salt concentra- andenjoys an appreciable solubility in waterthat tion,pH, temperature and the presence or absence facilitatespost-depositional growth and textural oforganic/biologicalmoleculesdetermine changes.The efŽ cacy of the silica deposition whetherthe coagulum gels or develops as a processand the strength of the residual bonds that solidphase (e.g. Iler, 1979, pp. 364 –407). linkthe microspheres is re ected in the rapid Typically,silica colloids are <5 nm in diameter. buildup of silica on all available surfaces Asthey aggregat e,the radius of curvatur e throughoutthe Wairake idrain,despite the increasesand the surface of the growing particle comparativehigh speed and turbulence of the Žllswhere small colloid particles adhere to a dischargewater. surfacerather than one another. It is an extension Thevariations observed in the size and shape of ofthis process that can produce the microsphere theX-rayscattering broadband are consistent with structureand nanosphere substructure of the establishedmodels of behaviour in young non- juvenileWairakei silica strands. crystallinesilicas. When Ž rstremoved from its Allsamples examined show a limitedrange in parent uid,opaline silica loses water and its particlesize and shape. Once silica particles grow juvenile,open structure collapses (Iler, 1979).

OK WK = Wairakei drain 1.40 OK WK OK = Orakei Korako sinter terraces WK 1.35 WK ) WK

Å WK WK ( WK WK d WK 1.30 OK OK WK OK

1.25 OK 0.1 1.0 10 Age (y)

FIG.10. Extension of the sinter aging model of Herdianita et al. (2000a)to include data from the present study. Plotted points bear error bars showing both possible age ranges and standard errors for FWHM.

577 B.Y. SMITHETAL.

Presumablyit isthischange that is re ected in the Blackall, L.L.,Hayward, A.C.,Sly, L.I.(1985) initialreduction in FWQM, FWHM andFWTM Cellulolytic and dextranolytic Gram-negative bacter- andthechangesinscatteringintensity. ia: revival of the genus Cellvibrio. Journal of Subsequently,as thescattering intensity increases Applied Bacteriology , 59, 81 –97. andthe half widths decrease further, if, as in Blank, C.E., Pace, N.R.and Cady, S.L.(1999) A clays,the half width is taken as a measureof molecular biogeographic comparison of microbial latticedisorder, then a smallreduction in the communities in near boiling silica-depo siting disorderof theopaline structure is implied during Yellowstone thermal springs; relating communities theearliest maturation of the deposit. Certainly, to siliceous sinter textures .Geological Society of thehalf width changes observed in the present America Annual Meeting,Denver Colorado, studyare consistent with the pattern of behaviour Abstracts with Programs 31(7), 325. foundby Herdianita et al. (2000a)whoreported Blatt, H., Middleton, G.and Murray, R.(1980) Origins of Sedimentary Rocks 2nd edition. Prentice-Hall, anoverall decrease in FWHM withtime. The NewJersey, USA. resultspresented here extend the range of their Brosius, J., Dull, T.L., Sleeter, D.D.and Noller, H.F. agingmodel, plotting in the lower part of the (1981) Gene organisation and primary structure of a rangeof their year-1 values (Fig. 10). Whatever, ribosomal RNAoperon from Escherichia coli . theprecise nature of the processes involved, the Journal of Molecular Biology , 148, 107 –127. evidencepoints to substantial post-depositional Cady, S.L.and Farmer, J.D. (1996) Fossilization structuraland textural changes occurring at avery processes in siliceous thermal springs: trends in earlystage in opaline silica. Hence, samples preservation along thermal gradients. Pp. 150 –173 collectedfrom extinct hot springs or evenmodern in: Evolution of Hydrothermal Ecosystems on Earth depositingsprings have experienced substantial (and Mars?). Ciba Foundation Symposium, Wiley, texturalandsilica phase modiŽ cation. The Chichester, UK. evidencesuggests that even specimens stored in Farmer, J.D. (2000) Hydrothermal systems: doorways to museumcollections have been subject to an early biosphere evolution. GSA Today, 10, 1 –9. indeterminateamount of change and questions Herdianita, N.R., Browne, P.R.L.,Rodgers, K.A. and theextent to whichtheir mineralogy and textures Campbell, K.A. (2000 a)Mineralogical and morpho- mightbe used to elucidate details of the original logical changes accompanying aging of siliceous depositionalprocess or, perhaps, the speciŽ cs of sinter and silica residue. Mineralium Deposita , 35, themicrobes involved. 48 –62. Herdianita, N.R., Rodgers, K.A.and Browne, P.R.L. (2000b)Routine procedures for characterising Acknowledgements modernand ancient silica sinter deposits. Thanksare due to Contact Energy for allowing Geothermics , 29, 65 –81. accessto and sampling of the Wairakei Power Holt, J.G., Krieg, N.R., Sneath, P.H.A., Staley, J.T. and Williams, S.T.(editors) (1993) Bergey’s Manual of Stationdischarge drains and to Ruth Smith and th Lew Baconfor supplying additional information. Determinative Bacteriology 9 edition. Williams Phillipaand Craig Gibson granted access to and and Wilkins, Maryland, USA,787 pp. allowedsampling of the sinter terraces at Orakei Hugenholtz, P., Pitulle, C., Hershberger, K.L.and Pace, N.R.(1998) Novel division level bacterial diversity Korako.KAR receivedfundingfromthe inaYellowstone hot spring. Applied and Universityof Auckland Research Committee Environmental Microbiology , 180, 366 –376. andBYS fromEnvironment Waikato and the Iler, R.K.(1979) The Chemistry of Silica: Solubility, JeffAllen Memorial Fund of the University of Polymerization, Colloid and Surface Properties, and Auckland. Biochemistry. Wiley-Interscience, NewYork, 866 pp. References Inagaki, F., Hayashi, S.,Doi, K., Motomura, Y., Izawa, E.and Ogata, S.(1997) Microbial participation in the Altschul, S.F.,Gish, W.,Miller, W., Myers, E.W. and formation of siliceous deposits from geothermal Lipman, D.J. (1990) Basic local alignment search water and analysis of the extremely thermophilic tool. Journal of Molecular Biology , 215, 403 –410. bacterial community. FEMSMicrobiology Ecology , Amann, R.I., Ludwig, W.and Schleifer, K.H.(1995) 24, 41 –48. Phylogenetic identiŽcation and in situ detection of Inagaki, F., Yokoyama, T., Doi, K.,Izawa, E.and Ogata, individual microbial cells without cultivation. S.(1998) Bio-deposition of amorphous silica by an Microbiological Reviews , 59, 143 –169. extremely thermophilic bacterium Thermus spp.

578 OPAL-A AND ASSOCIATED MICROBES

Bioscience, Biotechnology and Biochemistry , 62, Cloning: ALaboratory Manual , 1-3, 3rd edition. 1271 –1272. Cold Spring Harbor Laboratory Press, NewYork. Inagaki, F., Motomura, Y., Doi, K., Taguchi, S., Izawa, Smith, D.K.(1998) Opal, cristobalite, and tridymite: E., Lowe, D.R.and Ogata, S.(2001) SiliciŽed noncrystallinity versus crystallinity, nomenclature of microbial community at Steep Cone Hot Spring, the silica minerals and bibliography. Powder YellowstoneNationalPark. Microbesand Diffraction , 13, 2 –19. Environments , 16, 125 –130. Sunna, A.and Bergquist, P.L(2001) Agene encoding a Landmesser, M.(1995) Mobilita¨tdurch Metastabilita¨t: novel extremely thermostable 1,4-beta-xylanase SiO2 Transport und Akkumulation bei niedrigen isolated from an environmental DNAsample from Temperaturen. Chemie der Erde , 55, 149 –176. Kuirau Park, Rotorua, NewZealand . GenBank Lloyd, E.F.(1972) Geology and hot springs of Accession number AF402975. Orakeikorako. NewZealand Geological Bulletin , Weed, W.H. (1889 a)On the formation of siliceous 85, 1 –164. sinter by vegetation of thermal springs, American Miller, D.N., Bryant, J.E., Madsen, E.L.and Ghiorse, Journal of Science , 37, 351 –359. W.C. (1999) Evaluation and optimization of DNA Weed, W.H.(1889 b)Formation of travertine and extraction and puriŽcation procedures for soil and siliceous sinter by the vegetation of hot springs. US sediment samples. Applied and Environmental Geological Survey 9th Annual Report, 1887 –1888, Microbiology , 65, 4715 –4724. 613 –676. Rodgers, K.A.(2000) Research on silica sinters. Mineralogical Society Bulletin , 126, 16 –17. [Manuscript received 8August 2002: Sambrook, J. and Russell, D.W. (2001) Molecular revised 6February 2003 ]

579