Natural Bacterial Assemblages in Deep-Sea Sediments: Towards A Global View

NATURAL BACTERIAL ASSEMBLAGESIN DEEP.SEA SEDIMENTS: TOWARDS A GLOBAL VIEW

J.w. DEMING and P.L. YAGER Schoolof OceanographyWB-|0 Univer sityof Washington Seattle,Washington98195 USA

ABSTRACT. Deep-seasedimentsrepresenta significant sink term in the global carbon budget. Environmentalconfols on the residentbactcrialassemblages, which determinein largepart the magnitude of that term, are only poorly understood. In searchof the primary constrainingfactors,we examined existingdataon bacterialbiomassin sedimentsacrosstie entiredepth rangeof dre oceans,along with a smaller data set from abyssalNorth Atlantic and Arctic basins. The latter included measuresof bacterial biomass,utilizationratesof dissolvedamino acids,and the supply of particulateorganiccarbonto the seafloor. A statistical analysis of these data indicated ocean depth to be a weak predictor of bacterial pilrameters,while the magnitudeof organic carbon flux to the seafloorproved a strong one. We reached essentiallytwo conclusions:1) the missingfactor in our undcrstanding of bacterialactiviticsin deep-sea sedimentsis the quality or hydrolytic potential of the organic carbon supply to the seafloor; and 2) highlatitude basins,from which the highestbenthic bacterialbiomassand activity rates were recorded,should provide ideal sitesfor testinghypotheseson the bacterialfate oforganic carbonin the deepsea.

l. Introduction Global pattemsof abundance and distributiono[ macrofaunaand mciofaunaon the seafloorare (Rowe, 1983;Thiel, 1983). Early Sovietwork (reviewedby Rowc, 1983) wcll established suggestedthat dcnsityof benthicfaunadccreasedwith oceandepth,distancefrom shore,and decreasinglatitudcin relationto one crucialfactor,the availabilityof food resources.Modem studiesin which the flux of particulateorganiccarbon(POC)to the benthoshasbeenmeasured directly(e.g.,Tietjenet a1.,1989;Rowe et al., 1991)haveonly strengthened this view. No comparableunderstanding of bcnthic nanobiota(2 to 50 pm size range)has beenpossibleto achieve,due to a vcry limited database. Progresstowardsa global view of seafloorbacteria, however,is proceedingrapidly. In a recentrevicw,DcmingandBaross(in prcss)culledfrom the literaturcover 650 countsof bacteriain surfacc(0-1 cm) sediments to dcvelopseafloorabundance anddistributionpattemson a global scale. Bccausethe greatestnumberof bacteriaon an occan-widebasiswas shownto occurat abyssaldepths,thc authorsthenexaminedmorecloselya suiteofdata from five deep-sea locationsin searchof potcntialenvironmental dctcrminant(s).In keepingwith our understanding of benlhic fauna in gcncral,total bactcrialbiomasswas found to correlateinversely,though poorly, with occandepth,somewhatpositivclywith total organiccarbonin the sediments,but moststrongly(andpositively)with the flux of POCto theseafloor. l1 G. T. Rowe and V. Pariente (eds.)Deep-SeaFood Chains and the Clobal Carbon Cycle, ll-27. @ 1992Kluu,er Acadenic Publishers.Printed in the Netherlands.

.l

Hr'rc \IL-c-\pandthc globalsct of bactcrialabundancc datacxaminedby Dcming and Baross rn prl'ss'toincludcthoscrcportcdby Thicl ct al. (1989tfromtheabyssal NorthAtlantic,aswcll a-rprr-liminaryr'aluesfrom dccp Arctic basins(Dcming, unpublished)that allow a first 3-i1'ssmcnl of latitudinalinflucnces.We supplcmcnt(andconfirm)the pattemsof maximal ba;tr'nalgro*ti ratcson thc dccp seafloorprcscntedby Deming and Baross(in press)by rclatcdmeasurcmcnts .crLirdc'nng obtaincdbl LochtcandTurley(1988)andTurlcy andLochte l99Or In an auemptto differentiatc bcnthicbacterialactivityfrom simplcmeasurements of rr.rndingslor-k.*c aJsoprcscntand examincratcsof bactcrialutilizationol dissolvedorganic ;:tbon tDOC: spccilicalll.aminoacids)in scdimcntsof broadgcographicdistributionin the \.-rnh .{tlanticrfmm Dcmingand Col*cll. l9E5: Roqe and Deming,1985;Wilke et al., 1985; r n , J D c - m i n gu . n p u b l i s h c d ti.n c l u d i n gt h c G r c c n l a n da n d N o r w c g i a nB a s i n s ( D e m i n g , r Finalll. \ar-projcctbiologicalandchcmicalconstraints unpublrshr'd on abyssobcnthic bacteria th3tm3\ prorc crucialto a prcdictivcundcrstlndrng of thc rolc of dccp-sca bactcriain recycling organiccertrcnat [h!-seafloor. l . \ I a t e r i a l sa n d \ t e t h o d s

Thc bacterialdatathatr"c cxaminein this papcrarc dcscribcdin Lhcscctionsthat follow,along uith thc mcthodsthat wereuscdto gcncratcthcm. Wc sclccteddatafrom scdimentsamplcs (e.g.,RowcandDcming,1985)or multiplccorcrs(Bamettet rccovcrcd in USNEL-typcboxcorcs a l . .1 9 8 - 1T:h i c lc t a l . ,1 9 8 9 ) .I n a l l c a s c s , t h e c o r c s h a d b c c n k c p t c o l d u p o n r e c o v c r y a n d , a s s o as possiblc,sampleswerc eitherfixed for cpifluorescence microscopyor processed for rate dctcrminations undersimulatcdin situtempcraturc andpressure.Measurcmcnts of POC flux to thc scafloor,usingcylindricalsedimcnttraps(3:1 aspcctratio)moored10-92m abovethebottom, wcrealsotakcnfrom thc literature(Roweard Dcming,1985;Wilke ct a1.,1985;Tietjcnet a1., 1989;Roweet al., 1991)and from recent,comparablc work in high-latitudc basins(Rowcand Dcming,unpublishcd). 2.1 SAMPLING LOCATIONS Generalinformationaboutthc samplinglocationsthat form the basisof most of our analysesis prescntcdin Table 1. Thc stationsspanthc North Atlantic, from as far southas the Demerara AbyssalPlainncarthe Amazonconc(t1.5'N)to thc northemmost sitein the GrccnlandBasinnear (75"N). For six of theselocations(markcdby asteriskin Tablc l ), a comparable Spitsbcrgen suitc of informationon bacterialbiomass,DOC utilizationrate,andPOC flux ratewas available,thus enablingthe calculationof Pcarson'sproduct-rnoment and partial corrclationcoefficicnts accordingto Sokaland Roh-lf(1981;scc Scclion3.4). We intentionallyomitreddatafrom a scventhsitc (HattcrasBasin)duc to unccrtainticsaboutthc scdimenttrap results(Wilke et al., 1985). Includingthe data from this sitc would acccntuatcdifferencesbetwcenhigh- and lowlatitudedatadiscusscd bclow,but wouldnot changeour gcncralconclusions or prcdictions. 2.2 BACTERIALABUNDANCE AND BIOMASS To devclopa databaseon thc apparcnt"bactcrialcarryingcapacity"of the scabcd,we selccted abundance measuremcnts that allowcdintcgrationto a scdimcntdcpthof 15 cm andcxtrapolated thevalucsto an arealbasisof squaremctcrs(RowcandDeming,1985;Tictjcnct al., 1989;Rowe ct al., 1991;Dcmingand Baross,in prcss;Dcming,unpublishcd).Sincein no known casedo

l_)

bactcriacci*- to erist in scdimcntsbelowa depthof 15 cm, eachvalueis a conservative cstimate of seabcdden-6000 a D -

d

560-

345

f

20'

186 g

7.lD

h

032J

21.6i

96

I

t.v

J

Dcterminedby epifluorescence microscopy. SeeDemingand Baross(in press)for all citationsexamined.

AustraliancoralreefsedimenLdeprhof 5 m (Hansenet al., l9g7). " Ausrralianmangrovesedimcnt(Alongi, l9g7a). e Califomiahydrocarbonseepsedimcnt,depthof lg m (Montagnaet ai., 1989). fKiel Bight muddysedimenr, dcpthof 28 m (Meyer-Reil,l9g7a). I Australianreef slopescdiment,depthof 695 m (Alongi, l9g7b). h " Narcsabyssalplain sediment,dcpthof 5840m (Demingand Baross, . in press). ' Nova Scotiancontinenhlrise sedimcnt(HEBBLE site),depthof . 4626m (Thistleer at., 1985). J Pu.rto Rico Trenchsediment,depthof 7460m (Tietjencr aI., l9g9).

Whcn totalbacterialbiomassto a subsurlacc dcpthof 15 cm is considercd, thc relationship bctwcenapparcntbactcrialcarryingcapacityof thc scabedandoccandcpth appcarsas in Fig. 2. Fcwcrdatapointswcrcavailablcfor thisanalysis (n = l8 vcrsusn > o5b tor surfaccscdimc=nts) with thc rcsultthat thc high cnd of thc biomassscalcin thc figut(at a shallowsite)appeari undcr-rcprcscntcd'Howcvcr,biomassmcasurcmcnts in rhc surfacelaycr alone from most shallowcnvironmcnts (rcvicwcdby Mcycr-Rcil,l9u7a;DcmingandBaross, in prcss)arcashigh asthatindicatcdin thc ligurc.

G l o b a bl a c t e r i aNl o . ( l o g r o) in surfacesediments(0-1cm) 27

zo

Inle!'loarano St,Otldal ( 6000m)

in thesurfacelayer(0-1cm) of marine Figurel. Globalestimates of bacterialnumbers fromrangesgivenin in press),calculatcd (modifiedfrom DemingandBar"oss, sediments for andusingarealcoverages cm-r whennecessary Table2, assuming 1.5g DW sediment (1942). thedepthzonesreportedin Sverdrup 3.2 BACTERIALGROWTHRATESAT THESEAFLOOR in reasmeasured Maximal growthratesof aerobichctcrotrophicbacteriain dccp-seasedimcnts, pressurized, DOC-enrichedslurriesof box-corcsurfacesedimcnts,appearto be slow and growth ratesof relativelyinvariantwith oceandcpth,as shownin Fig, 3. Similarly-dctcrmined with samplcsol "frcsh" sedimcnts,rccoveredin short-termbottom-moored bacteriaassociated sedimcnttraps,diminish with increasingdepthbut ncver to a lcvel as low as that measuredin underlyingsedimcnts(Fig. 3). Ratesmcasurcdin dccp-seasamplesenrichedwith naturallevels of phytodetritusfrom thc surfaccof a multiplc corcr (Lochtc and Turley, 1988:Turley and trap samples(Fig.3). Lochte,1990)fall within thc rangcof thoscmcasuredin the DOC-enriched

SEDIMENTS 3.3 RATESOFDOCUTILIZATIONIN DEEP.SEA shouldbc avoidcd. This In general,inferenceof biologicalactivityfrom biomassmeasurcmcnts is particularlytrue for bactcria,sincc that portion of the total populationactive on a givcn

a a (o^ E': E

zr

-:v v i; or

4000

3000

tiF

CD

2000

1000

0 0123456789

Depth(km) Figurc2. Logarithmicdccreascin benthicbacterialbiomass(mg C m-2 to a scdimentdcpthof l5 cm) asa functionof oceandepth(km). Thc solid line resultsfrom parametriclinearregrcssionpcrformedon semifog transformeddara (r2 = .739). Sce Table 1 for symbol kcy to study siG informationand literature citations.

substrate at a givcn time is almostalwaysunknownandunpredictable.Doesa vcry activebut small portion,or a lessactivebut largeportion,accountfor thc overallbulk ratc measurement? With this caveatin mind, it nevertheless seemsnotcworthythat the correlationbetweenbacterial DOC utilizationratcsand parallelbiomassmeasurcments over an abyssaldepthrangeof 37005840m in the North Atlanticis significantandstronglypositive(r = 0.925;n = 7; Fig. 4). 3.4 STATISTICAL ANALYSES In an attemptto bctterundcrstand the environmental importanceof oceandepthand pOC flux to bacterialparamctcrsin the decp sca,we calculatcdPearson's product-momentand partial correlationcocfficients(SokalandRotrll 1981)for datafrom six North Atlanticlocations(Tablel). Thc data are listcd in Table 3. Product-momcntcorrelationcoefficients(rxy ) and partial correlationcoefficicnts,independcnt ofone (r*r.2 ) andthcn two variables(rxy.rw) areshownin Table4. From thcscstatisticalanalyses, depthemcrgcsaslhe weakestdeterminant: in eachcase, pairedparametcrs covaryindcpendently ofoceandcpth(Table4, footnoteb). On theotherhand,

='

E o =

2.O

r<

o (E

=

oo o-

1.0

'= o C) d

.o

nc

o q)

Depth(km) in bacterialpopulations growthratesof aerobicheterotrophic Figure3. Comparative 10-92 m deployed traps sediment (recovered in short-term "fresh" particulates of samples aboveseafloor,from Demingand Baloss,in press,op€ncircles),includingphytodetritus recoveredfrom the seabed(Lochteand Turley, 1988,opentriangle;Turleyand Lochte, slurries(box cores,from Demingand of sediment 1990,opensquares), and in samples Baross, in press,solidcircles).Usedepthsaskey to sitelocationin Table1.

POC flux to the seaflooris shown to influenceboth bacterialparametersstrongly (Table 4. footnotea). This result agreeswith the paradigmfor larger size classesof the benthos,that particulatefood is the critical dcterminantof biomass. Other less readily interpretedresultsfrom below,includingthc theseanalysesthatpoint to featurcsuniqueto benthicbacteriaarediscussed with occandepthwhenthe effectsof both POC flur implicationttratbacterialbiomassincreases andDOC utilizationarercmoved(Table4, footnotec). AT HIGHLATITUDES 3.5 OBSERVATIONS Table3 andFig. 5 revcalthatthe highestbacterialbiomassandDOC utilizationratesin deep-sea sedimentsderivefrom thehighestlatitudesyet studiedby thesemethods. A furtherobservation

4.5

40

N^

,. ql v;t.t

o6 :

z.o E

i( l.u

m

nt

0.0 0

s0

100 150 200 2so 300 350 400 450

Bacterialbiomass (mgC m-2) Figure 4. Linear regressionanalysis of aerobic bacterial rates of DOC 1laC-amino acid) utilizationand bacterialbiomass,eachintegratedto a sedimentdepthof 15 cm (r2= .855; symbolkey in Table l).

TABLE 3. Data usedin statisticalanalyses(Iable 4) and in Figs. 5 and 6; orderedaccording to increasingbacterialbiomass. (Use oceandepth as key to additional information in Table l).

Bacterial biomass (mgC m-2)

DOC util. rate (mg C m-26-1;

16 82 116 210

0' 1.70

4", A

2.91 4.02

449 u

T*

r.23 2.05

POC flux , (mg C m-'d-') 1.45 6.00 9.50 14.8 11.5 21.5

Ocean depth rumi 5.8 4.1 4.8 4.4 5.t ).t

Station latitude("N)

23.O 47.5 10.5 8.5 75.0 66.5

lo* to detectduring the 2-d experimentalincubationperiod.

apparentlyunique to high-latitudebasinsis the significantlyincreasedefficiency with which benthic bactcriaconsumeDOC. The typical percentageof laC-amino acids partitionedto respirationby aerobicbacteriain re-pressurized sedimcntsfrom temperateregionsof the North Atlantichasalwaysaveragedabout90Vo(Demingand Colwell, 1985;Rowe and Deming,1985; Wilke et al., 1985;Deming,unpublished).Very little of the substratewas found to fuel

from the samples ucre appliedto sedimcnt of newbiomass.Whcnidcnricalmcrhods production signifiCantly decrease to found was pcrccntagc Grecnlandand NorweglanBasins,the rcspirafory in termsof percent arepresented obsen-ations rDcming,unpubtishcd),In Fig. 6, thesccontrasting into biomassaccordingto stationlatitude' DOC incorporatcd coefficients (r*, ) andPartialcorrclation coefficicnts correlation product-moment TABLE-1.Pearson's = (n 6)' 3 fromdatain Table ,r,.r., ild ,*t..r*) calculatcrl

t*)

\. \

D-pth. Broml.s

'*y.t of a singlcvariable) (Independent

,9s0

- l-i-1

I 60a

.943c

.'7664

-.976

fI-nth

DOC rat.^tupth. POC llux

POCflux. Biomass POCflux, DOC rate Biomass, DOC ratc

-.916 -.816

.980 < iA

.945

.941

-.175

.378 .867b

450

.939

rxy.zw (Independent of both)

.528

.4174

.855b

-.058

.980b

.955

for theconelationbetwcenx andy' significantly a Low r*r., valuesindicatethatPOCflux accounts x andy weakly,if at all' between thcconelation UHigtti-r, valucsindicate rhatdepthinfluences whcn effectsof DOCrateand depth with ocean increascs c Hiln r*i.r* valueimplicsthatbiomass POCfluxarercmoved' levelof o,< 0.05. significance Boldprintindicates

4. Discussion of the our biologicalunderstanding In a rccentreview,GoodayandTurley (1990)summariz.ed of benthic biomass and abundancc the which in environment "a food-limiied as dccp sca This relation organismsis directlyrelatedto the amountof footl reachingthe scdimentsurface' (Deming Wc confirm megafauna." up"pti.,to the mcioiauna,macrofauna,and deposit-fccding It is also list' to this added now be can bacteria press) that in uni Bu.orr, in press;Roweet a1., constitute sediments in dcep-sca bacteria densitics, low spccific rclatively of clearthat, in spitq theirpotential the largestglobalfractionoftoial benthicbactcria(Fig. l, Tablc2). Understanding is critical to 1990), (Jahnke et al'' and actualratcsof activity, especiallynear oceanmargins prcdictionsof globalcarbonflux at the scafloor.

o U)

g^ F

C\I

.c 0 'oE

300

t)

,s 's cD 200 o

0 4.5 4.0 .o Nj-

=,9.v

Eo

On OE

O rr

3.5 en

2.5 2.0

E g 15 H

1.0 0.5 0.0 6

I

101214

16182022

VerticalPOCflux ( m gC m - 2d - 1 ) Figure 5. Linear regressionanalysesof bacterial biomAss(A) and DOC utilization rate (B) versusPOC flux (symbol key in Table 1, data from Table 3).

4.I BACTERIALBIOMASS Across ocean dcpth zoncs, the logarithmic decreascof bacterial biomass indicates an apparenl reduction in the carrying capacity of marine scdiments (Table 2, Fig. 2). Bcyond contincntal margins, howevcr, biomass docs not decreasefurthcr with depth to a significant degree. Thus,

80 7i

bU

o

(n

() ;.

40 an

@

20 10

Latitude("N;

into biomassaccordingto (l4C-aminoacids)incorporated substrate Percent 6 Figure Table1)' sutionlatitude(symbolkeyin

The biomasson the deep seafloor' depth are controlling bacterial reproduction dcrcrminanrsothcr than ocean bacterial of rates ."-tmal p.".rur" may set . i.p "t cffccrs of elevatedtryorosiatlc observedtrend in bacterialbiomass fu.g.iut'tte in e*ptains rFig. 3), but the ,"ppty oir"Ot rTablc-1,Fig 5). Wc|ullySuspectthatthequalityorlabilityo_frhePoCfluxisalsocriticaltobenthicbacteria and distancefrom photosynthetic occursou.ing t.ansii, in" ti*" biomass. Sincedcgradation We caninfer an effecton biomass rf,. pOC reaihingin. t".n""t. oi quuf,r' thc affect will source measuredat a quiescentsite ,,end-members,,i;3il;; .^atin"O Lo*..t biomasswas Sea' Sargasso from poc: trr. N.r.rluvrrui pruinunderlyingthe "t primlry of levels lcasrrikcrr ro rccci'c p";;r";;;;;h high bY i'o* utt'i' 'itt' innutntJ seasonally in press:me Hiehcstbiomass*u' '"toioJ vcnicat mixing (Smithet al" an9bv-gf;ic tht .t prJductivitl analysesof the Poc' however' ""ffi;;-*lt drguni. .t.rirur o. hydrolytic xor*.cgran's.ur. and Grecnland wittibenthicbacterialmcasurements' statistically'benthic rcmainro bc obtaincdin tontt't DoC utilization rate are removed The simplest \\.hcn rhc .rr..r, oiioc flux and it.t9t",1:.lYnott c)' increasewitt, oceariffi bacrerialbiomassupp"urcto c r p l a n a t i o n f o r t h i s . , . n o i . t h a t p r e d a t i o n p , . , *no*e , " o n uet e nal' t t r (in i c b press)' a c t e r i awho d i mfound i n i s h ethat s w ithe thocea and ro"t'it-t*iu;;;;;;"'"t;111m depth with Support dcpth. bacteriaincreased a ,,total"benthic biomassuu.iiuiuur" or complementary perccnragcof rhc sites. 90% .*ty.ttr .An,altemate Na*ral bacterial distanccfrom shore,.ppr".iiri"g "t*ost sea' deep qr.rt,v of food r#urces in the -u..-r.q,i*O (Deming compounds cxplanation.guin ,nuotk'.-r-in. tot' ior ttreAeg.aiaiionof 'ir,t. :tl"t-:* .on*n,u depth' o, ocean asscmblagcs supptyincreaseswith ,.rru.rory portion o?*theroc prcss). utilize to in Baross. necessary and may be ;;i;e?) bcnthicUu'.t.,iJ tot*unltitt rhcnrclativcll,targer(";'r;;. ir.

-{.2 BACTERIALACTIVITY Availablemcasurements of bacterialactivityin deep-sea scdimentsare problematicif one is seekinginformationon in situ ratcs. We have cxaminedtwo measuresof aerobicactivity: 1) maximalgrowth ratesinfluencedby both sampledilution and organicenrichment;and 2) DOC utilizationrates thatmay overestim atein situ ratcsdueto sedimentdilutionbut not to enrichment. At this point, we havebeenunablcto identify any significanttrendsof the growth rateswith environmentaldeterminants suchas POC flux (analysisnot shown). However,growth ratesin sedimentsamples,rcgardless of oceandepth or latitude,were uniformly lower than those measuredin collectionsof particulatematter recentlyarrivedat the seafloor(Fig. 3). A copiotrophicversusoligotrophicnaturc of the dominantbactcriain "fresh" versus"buried" sediments canbe invokedto explainthis pattem(Deming,1986),but it may alsobe a reflectionof differentialdilution effccts. Bacterialand particledcnsitiesin the sedimenttrap sampleswere aiwayslower than in thc scdimentslurrics. Other factorsinfluentialto bacterialgrowth rates, such as bacterivoryby protozoans(Turley and Lochte, 1990), require further experimental investigation. In contrastto the growth rates,the aerobicDOC utilization rateswere obscrvedto vary significantlywith POCflux (Fig.5B). This correlationappearsto be coincidental, howcver,since it canbe explainedpartiallyby the effectof biomassandcompletelyby the combinedinfluenceof biomassand depth(Tablc4). That DOC ratedecreascs significantlywith occandepth(Table4) may againindicatea capimposedby hydrostaticprcssure. The strongcorrelationbctweenaerobicDOC utilizationratesandtotalbacterialbiomass(Fig. 4) was unexpected. An unknown and presumablyvariablc fraction of the natural bacterial assemblage in marinesedimcntsis believcdto be inactive. In sedimentswhere anacrobic proccssesfigure significantlyin the degradationof organiccompounds,mcasuresof aerobic activitywould not bc cxpectedto corrclatewell, if at all, with the total standingstockof bacteria. Our statisticalanalyses suggestthata directcff'ectof total biomasson aerobicDOC consumption ratcsmay bc moremcaningfuland predictablein the oxygenateddccp seathanelsewhcrein thc ocean. For an individualorganism,an increasedratc of DOC utilizationwill not translateinto increasedbiomassunlessa significantfraction of thc cnergy uscd is partitionedto anabolic functionsrathcrthanrespiration.We haveobservcdincorporationefficicncyto relateto latitude (Fig. 6), with highestperccntagesrecordedfor bacterialpopulationsin rhe polar sedimenrs cxamincd. For high cfficienciesto explainthe high biomassand DOC ratesalso observedin thesescdiments,the fractionof the bactcrialpopulationsthat is activemust be smallcrthan in low-latitudcsedimcnts or our calculations of biomassfrom bactcrialnumbcrsusing a singlc conversionfactormust be inappropriate.To betterdcciphercause(s)underlyingtheseelevated incorporationelficienciesin polar sediments, furthcrstudyof potentiallatitudinaldifferencesin bacterialsize,communitystructure,and food quality are needed. Eltccts of thc colder(-l'C) temperaturcs charactcristic of thcsesediments mustalsobc considered. 4.3 ANIMALEFFECTS Simplisticallyrclatingbactcrialparametcrs to mcasurcsof organiccarbonsupplyor consumption in deep-sca sediments appeanto overlookthc effectsof foragingactivitiesofbcnthic faunain thc samc environmcnt. Such cffects are bcyond thc scopeof this discussion,but thosc begging rescarchin rclationto bacterialactivity (Plantcct al., 1990; Demingand Baross,in prcss)

/+

chambcrsranimalgut-s'fccal *r-llcts)for improved include:1) the provisionof fermentation bacrcrialconsonis:and 2) the intermittent anacrobic by degradationof refractorycompounds by constrained acti\it\. into scdiments of acrobic Stimulation consequent iryiction of oxygcn,and and egcsted ingcstcd, in occur actionsof bactcria diffusion. Howevcr,sincethc degradative ratcsb1 different bacteria: faunally undisturbedsedimentsalike (albcit at potcntialll altcrc'd wc expect in prcparation)' Dcming. and Vcttcr 1990; al., ct oeming and Colwcll,1982;Plante fatc of organic prL-dict thc to provc sufficicnt will ultimately paramcters that selectedbacterial sedimcntbacteriawith carbonin dcep-seasediments.The incrcasedrelativcbiomasslcvcls of supply(Roweand Poc of thc consumption oceandepth(Rowcet al., in press),their significant in Table3)' and ratc Doc and flux PoC of valucs comparative Dcming,1985;Smithct al., 1'987; sediments of dccp-sea characteristic compounds rcfractory of clegradation their dominancein thc to this idea' (DemingandBaross,in press)lendcredence ENZYMES 4.4 EXTRACELLULAR thatwc haveprcscntcddo not takcinto accountthc fact thatbactcriacannotuse Thc relationships cxtraccllularly'via POC directly. Largclabile or rcfractorycompoundsfirst mustbe hydrolyzcd For comparative mcmbranc' ccll thc across be transportcd can to a molccularsizethat enz.ymes, of thc carbon naturc to thc regards without mg c in in tcrms parameters all cast purpor.r, we have carbonpools nature of varied rccognizingthc ltiuing o. dead,particulatcor dissolvcd).However, activity and biological on and iays of pioccssingthcm is csscntialto retiningconstraints rcscarch' phasein dcep-sea thc predictivc reaching in deep-sea Towardsthis goal, we have initiatcd studicsof extraccllularenzymc activity (1989)to sluny (1986), Mayer and (1986), King Mcyer-Reil of mcthods the adapting sediments, andpressures.Our resultsarepreliminaryat andwholc-coreinjectionsundcrln slfu tempcratures from here. For example,sediments discussion thc to relevant this point, but a few of themsecm DOC utilization biomass, bacterial highcst the whcre Basins, Norwcgian the dcepGreenlandand rates'relativeto and POC llux ratcsin our datasetwerc observcd,also yieldedhigh hydrolysis in preparation)' Yagcr, (Deming and madein shallowcrenvironmcnts measurcments comparable activity)at (a of chitinase mcasurc ratc hydrolysis the glucosamine In fact,in surfaccsediments, substratc for any mcasured that was higherthan (370 atm) pressurc (-l'C) and ln,slratemperaturc 1986; Mcyer-Reil' rclcased; product of (based molcs on mcthod injcction by thc whole-core warrner much in a by slurry tcchniquc that measuredfor glucosaminc tbgf) anOapproachcd (protcaseandlipase) intertidalenvironment(King, 1986). Maxima for othcr enzymeactivities It seemsclearthat an depths' subsurfacc werc dctcctcdin thcsehigh-latitudescdimentsat cxtracellular persistcnce' quality, substrate POC incoming intcgratcdapproachof measuring our understanding significantly advance would parametcrs bactcrial other .nri.n. activitics,and of the fatcof organiccarbonat thedccpscafloor' REMARKS 4.5 CONCLUDING acrossthe By examininga still small,but intcmally consistcnt,sct of datafrom dccp-scabasins to the POC of supply thc North Atlantic and into polar latitudes,we havebeenablc to identify seabed' in the activity extcnt, to somc of bactcrialbiomassand, scaflooras a driving determinant that wouldimprovcour ability to predictthe fate We havealsoprojectedareasfor futurcresearch of the quality of POC reachingthe seafloor' studics including of organiccaibonat the seafloor, and thc efficicncy with which bacteria enzymes' of extraccllular activity the hydrolysis

2.s incorporatethe productsof hydrolysis Arcticregions.have emergedas ideal sites for testing hypothesesreraredto 1c!.oittrese suulects.,rrrt potu. r"ttrngr?-rl .*p".r"o ro figure most

;ftT3[':J#rl,i!*uti'urrt

ln.''n-'' oiglobal warmine 'JE, tiJn."ororsuch .;.;;;ii

ACKNO*LEDGMENTS Support for preparation^1f_rhi: manuscriprwas provided by NSF throughgrantsDPP-8800401,'oce-ssrisjj and oCE-9144237.i;thank Michacl encouraging Rex for us to explorea statisticaltrcaftnentoJthe avaitableoa;;, p;".Jumars for assistance in thoseefforrs,anob.il,nower"i r"{.r"i"s'u *oro workshopthat encouraged us to examine our datain a moreglobalcontext,randslices iotwithstandins. References

^tilf,l"?. i:?iti],t;:Triar

producrivitv andmicrobiar biomass in tropical mansrove sedimcnrs,

Alongi' D' (1987b)The distribution and composirionof deep-sea microbenthos in a bathyal regionofthc wcstemCoralSea,Deep_Sea Rcs.34, 1245_1254. Alongi' D' (1990)Bacterialg:owq .ut.r, p.oor.,ion andestimarcs of detritalcarbonutilizationin deep-sea sedimenm, Deep_5.un"r. zi,'lil_iqa. _ Barnetr,p.R.o., watson, J. conncrry,o: (19g4)A multiprc corcr tbr taking virtually .and undisturbed samplcsfrom shelf, bathyal-o uuyrrutscdiments, br.-otogi.u Acta7,gg_40g. Biscave,p.E., Anderson,R F and o.rl, nr. i,lsg)_a,u"., and consriruenrs eastemunited Statesconrinenral ro rhe slopeandrise:SEEp-1,c"",. "i;il;;; st.rin.s. g, g-55_g04.

".il:I,#,,?ff.,Iiiii

r rrss+lMi.,obi.i **o andDNA;y;;;;;, inmarine sedimenrs,

growth in decp-sea sedimcnr rrapandboxcore "?:x:'J.y;!]?fl rt;""ar samples, Mar.Ecol. tT,ii;J;i strategics of barophiric bacreria in rhedecpocean, Microbiol. #3-t?,ut"t"gical

Deming'J'w' and Baross,J.A. (in prcss)The early diagenesis of organicmattcr:Bacterial

oie-i." c.o*.'i,ir,vo,6,ropics .'l#iil;Jl#;"ffT1,:J'"i; in Y;,"f:'(J'1, _ Deming'J'w' andColwell'

R'R' (1982)Barophilic bactcria associated withthedigcstive abyssar tractsof holorhurians, Appl.Environ.-ruricrooioi. ^ 44, 1222-1230. Deming,J W. andcolweri,R.R. rrsssrobs;;"uiion,or nuropHri.microbialactiviry in sampres particurares

ff::i;iliH,.i#lfJ;::red

rro,n tr,eDemeiara erfrruieruin, Appr. Enviion.

Gooday,AJ. and Turlcy, c.M. (r990) Responses by_be,nthic organismsto inpursof organic matcrialro rheoceanfloor: A review,ptitJr. i.unr. R. Soc.Lond.A 331,l lg-13g. Hanscn,J.A.,Alongi,D.M., Moriarr.y, o-l.ri. ."0 pollard,p.c. (19g7)The dynamicsof benthic microbiarcommunirics ar Daviesn..t ..nt.ur creat Barrier n""r iiJ'ieefs 6, 63-70. Jahnke'R'A', Reimers,C.E. and c.uu.n, o.s. i,lo; tn,.nrinrution oi-..cycringof organic matterat thc seafloor nearocean margins,Nafure34g,50_54. Jannasch, H'w. and wirsen, c.o. (r9g3) rnlirrouiorogy of the deepsea,in G.T. Rowe (ed.), Sea,Vol. 8, Dcep_Sea The Biology,Wley_y",1.r.1.n..,fV.* york, pp. 231_259. .'X?f

activity i,i'r"-,.niJ.r marine sedimenrs, *"lfj'^f,#ff:H*;11;;o--Jr;;o,jou,.

_h

Lochtc,K. andTurley,C.M. (1988)Bacteriaandcyanobacteria associated with phytodetritusin thedecpsca,Nature333,67-69. \1a1er,L.M. (1989)Extracellularproteolyticenzymeactivity in sedimentsof an intertidal mudflat,Limnol. Oceanogr.34,973-981. \lc1cr-Rcil,L.-A. (1986)Measurement of hydrolyticactivityand incorporation of dissolved organicsubstrates in marinesediments, by microorganisms Mar. Ecol.Prog.Ser.31,143-149. \tc1cr-Reil,L.-A. (1987a)Biomassand activity of benthicbacteria,in M.J. Bowman,R.T. Barber,C.N.K. Mooers,and J.A. Raven(cds.),LectureNotes on Coastaland Estuarine Vol. XIII, Springer-Verlag, Studies, New York,pp.93-110. \1e1er-Rcil,L.-A. (1987b)Seasonal and spatialdistributionof extracellularenzymaticactivities and microbialincorporationof dissolvedorganicsubstrates in marine sediments,Appl. Environ.Microbiol.53, 1748-1755. \lontagna,P.A.,Bauer,J.E.,Hardin,D. andSpics,R.B.(1989)Verticaldistribution of microbial and meiofaunalpopulationsin scdimentsof a naturalcoastalhydrocarbonseep,J. Mar. Res. - 1 76. 5 7 - 6 8 0 . \ovitskl', J.A. (1987)Microbialgrowth ratesand biomassproductionin a marinesediment: Evidcncclor a very activebut mostlynongrowingcommunity,Appl. Environ.Microbiol. 53, )368-2372. Plantc.C.J., Jumars.,P.J. and Baross,J.A. (1990) Digestiveassociations bctweenmarine dctntivorcs andbactcria, Ann.Rev.Ecol.Syst.21,93-127. Rouc. G.T. (1983)Biomassandproductionof thc dccp-sea macrobenthos, in G.T. Rowe (ed.) Thc Sca,Vol. 8, Deep-Sea Biology,Wilcy-Intcrscience, New York,pp.97-122. Rouc, G.T. and Dcming,J.W. (1985)The role of bacteriain the tumover of organiccarbonin dccp-sca sedimenls, J. Mar.Res.43,925-95O. Ronc. G.T.,Sibuet,M., Dcming,J.W.,Khripounoff,A. andTietjen,J. (1991)Organiccarbon rcsidcncctimc in thedeep-sea benthos,Prog.Oceanogr.24,141-160. Roqe.G.T.,Sibuet,M., Deming,J.W.,Khripounoff,A., Tietjen,J., Macko,S. andTheroux,R. (in prcss). "Total" sedimcntbiomassand preliminaryestimatesof organiccarbonresidcnce timc in thcdeep-sea benthos, Mar.Ecol.Prog.Scr. S m i t h .K . L . , J r . , C a r l u c c i ,A . F . , J a h n k c ,R . A . a n d C r a v e n ,D . B . ( 1 9 8 7 )O r g a n i cc a r b o n mincralization in the SantaCatalinaBasin: Benthicboundarylayermetabolism, Deep-Sea R c s .3 - 1I.8 5 - 2 1 . Smith.W.O.,Codispoti, L.A., Nelson,D.M., Manlcy,T., Buskey,E.J.,Nicbaucr,H.J.andCota, G.F. (in prcss)Importanceof Phaeocystisbloomsin the high-latitudeoccancarboncycle, Naturc. Sokal,R.R.andRohlf,F.J.(1981)Biometry,W.H. FrccmanandCompany, New York. Svcrdrup, H.U.,Johnson, M.W. andFleming,R.H.(1942)The Oceans: TheirPhysics, Chemistry, andGcneralBiology,Prcntice-Hall, Inc.,New York. Thicl,H. (1983)Meiobenthos andnanobcnthos of thedeepsca,in G.T.Rowe(cd.)The Sea,Vol. 8. Dccp-ScaBiology,Wiley-Interscience, New York, pp. 167-230. Thicl. H., Pfannkuche, O., Schriever, G., Lochtc,K., Gooday,A.J.,Hcmlcbcn,C.H.,Manroura, R.F.G.,Turlcy,C.M., Patching,J.W.andRiemann,F. (1989)Phytodetritus on the deep-ssa lloor in a ccntraloccanicregionof thc NorthcastAtlantic,Biol. Oceanogr.6,203-239. Thistlc,D., Yingst,J.Y. and Fauchald,K. (1985)A dccp-seabcnthiccommunityexposedto strongncar-bottom currentson theScotianRisc (WestemAtlantic),Mar. Gcol.66,91-112.

Tictlcn.J H.. Deming,J.w., Rowe,G.T.,Macko,s.A. andwilke R.J.(19g9) , Mciobenrhos of rhe HattcrasAblssal Plain and PuertoRico Trench:Abundance,biomassand associations with bacreriaandparriculate fluxes,Deep-Sea Res.36, 156l-l5jj. Turlcl.C.\l andLochte,K.(1990)Microbialresponsetotheinp f ruetsohf d e t r i t u s r o t h e d e e p paleoecol,99,3_23. scabcd,Paleogeogr. Paleoclimat. Wilke,R.J..Deming,J.W.,Macko,S.,Tieqen,J.,Rowe,G., Stein,D., Fray, C., Khripounoff, A., Koshorek,D., Stepien,J. and voras,B. (19g5)Low level-waste oceandisposar project:n/v Isclin cruisel9 June- 10 July 1984,DataReport,BrookhavenNationalLaboratory Associated Unii'eniries, Inc.,Upton,New york. williams,P.M. andCarlucci,A.F. (1976)Bactcrialutilizationof organic matterin rhedeepsea, N a t u r e2 6 2 .8 1 0 - 8 1 l .