ebook img

Benthic communities in the deep Mediterranean Sea PDF

18 Pages·2013·1.3 MB·English
by  
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Benthic communities in the deep Mediterranean Sea

EGU Journal Logos (RGB) O O O p p p Advances in e Annales e Nonlinear Processes e n n n A A A Geosciencesc Geophysicaec in Geophysicsc c c c e e e s s s s s s Natural Hazards O Natural Hazards O p p e e and Earth System n A and Earth System n A c c c c Sciencese Sciencese s s s s Discussions Atmospheric O Atmospheric O p p e e n n Chemistry A Chemistry A c c c c and Physicse and Physicse s s s s Discussions Atmospheric O Atmospheric O p p e e n n Measurement A Measurement A c c c c Techniquese Techniquese s s s s Discussions Biogeosciences,10,4861–4878,2013 O O p p www.biogeosciences.net/10/4861/2013/ e Biogeosciencese n n doi:10.5194/bg-10-4861-2013 Biogeosciences A A c c c Discussions c ©Author(s)2013. CCAttribution3.0License. e e s s s s O O p Climate p Climate e e n n A of the Past A of the Pastc c c c e e s Discussionss s s Benthic communities in the deep Mediterranean Sea: exploring O O microbial and meiofaunal patterns in slope and basin ecosystepms Earth System p Earth System e e n n A Dynamics A Dynamicsc c K.Sevastou1,N.Lampadariou1,P.N.Polymenakou2,andA.Tselepides3 ce ce s Discussionss s s 1HellenicCentreforMarineResearch,InstituteofOceanography,P.O.Box2214,71003,Heraklion,Crete,Greece 2HellenicCentreforMarineResearch,InstituteofMarineBiologyandGenetics,P.O.Box2214,71003,Heraklion, Geoscientific Geoscientific Crete,Greece O O p p 3UniversityofPiraeus,DepartmentofMaritimeStudies,G.Lambraki21&Distomou,13I5n8s2t,rPuimraeeuns,tGarteieocne en Instrumentation en A A Methods andc Methods andc c c Correspondenceto:K.Sevastou([email protected]) e e Data Systemsss Data Systemsss Received:16November2012–PublishedinBiogeosciencesDiscuss.:11December2012 Discussions Revised:16May2013–Accepted:25May2013–Published:18July2013 O GeoscientificO p p Geoscientifice e n n A Model Development A Abstract.Thelong-heldperceptionofthedeepseaconsist- pendingontheMdaotadseetlu Dsede(vbaesleodponmtyepnetofcchabitat,ben- cc e e ing of monotonous slopes and uniform oceanic basins has thic component, taxonomic level). This, alongsswith the ob- Discussionsss over the decades given way to the idea of a complex sys- servedhighwithin-habitatvariabilitysuggeststhatotherfac- tem with wide habitat heterogeneity. Under the prism of a tors,whichtendtovaryHatyldocraollsocagley( haydnrodd yOnamics,sub- Hydrology and O highly diverse environment, a large dataset was used to de- strate structure, geochemistry, food quality, etpec.), may also pe scribe and compare spatial patterns of the dominant small- relate to the observed beEntahircthpa tSterynss.tOevmeran All, the results Earth Systemn A c c size components of deep-sea benthos, metazoan meiofauna presentedheresuggestthatdiffeSrecncieesnincsemsallce-sizebenthos Sciencesce s s andmicrobes,fromMediterraneanbasinsandslopes.Agrid between the basin and slope habitats are neithser strong nor s of 73 stations sampled at five geographical areas along the consistent; it appears that within-habitat variability is high, Discussions central-easternMediterraneanBasin(centralMediterranean, differencesamongdepthrangesareimportantaOndfurtherin- O p p northern Aegean Sea, Cretan Sea, Libyan Sea, eastern Lev- vestigationofpossibleenvironmentaldriversoefbenthicpat- Ocean Sciencee n n antine) spanning over 4km in depth revealed a high diver- ternsisneeded. Ocean Science A A cc Discussions cc sity, irrespective of the benthic group or level of taxonomic e e s s analysis. A common decreasing bathymetric trend was de- s s tected for meiobenthic abundance, major taxa diversity and nematodegenerarichness,butnodifferenceswerefoundbe- 1 Introduction O O p p tween the two habitats (basin vs slope). In contrast, micro- en Solid Earthen baniadltreicnhdnsetsosiinscsreiganseifiwcainthtlydehpitghh.eMr uatltitvhaeribaatesinaneaclyosseysst(eβm- C90ov%eroinfgthmeoorecetahnasn,6th0e%deoefSpthoseelaiedairs tEhth’saesrlutarhrfgaec Accesetaencdosnyesaterlmy Discussions Acce s s andδ-diversityandordinationanalysis)complementedthese on earth (Ramirez-Llodra et al., 2010). Yetsthe deep-sea s results and underlined the high within-habitat variability of floor remains largely unexplored in spite of the enormous benthic communities. Meiofaunal communities in particu- technological advances and the intense researOch efforts of O p p lar were found to change gradually and vary more towards the past few decades. The recent breakthrouenghs in deep- The Cryosphereen the abyss. On the other hand, microbial communities were sea research have neTvehrtehe lCesrsyreovseaplehdearceom Acplex system Ac c Discussions c highlyvariable,evenamongsamplesofthesamearea,habi- with diverse geological, physical and biochemesical charac- es s s tatandbathymetry.Asignificantproportionofthevariation teristicsthatsupportsawidevariabilityofdifferenthabitats of benthic communities and their descriptors was explained andbenthiccommunities. by depth and proxies of food availability (sedimentary pig- Thedeepseaconsistsofslopesandbasins.Theslopesare ments and organic content), but the combination of predic- thesteeppartofthecontinentalmarginsthatconnectthecon- tor variables and the strength of the relationship varied de- tinentalshelfwiththedeep-seaplains.Althoughofrestricted size(roughly10%,Ramirez-Llodraetal.,2010),slopesare PublishedbyCopernicusPublicationsonbehalfoftheEuropeanGeosciencesUnion. 4862 K.Sevastouetal.: BenthiccommunitiesinthedeepMediterraneanSea essential ecosystems for the functioning of the oceans and samples from Mediterranean slopes and basins, found that theglobe,astheyconstitutetheregionwherethecontinent- meiofaunal diversity in slopes was higher than in deep-sea to-oceantransferofwater,sedimentandenergytakesplace. plains.ThisdiffersfromtheresultsreportedbyVanreuselet Thesharpdepthgradientofslopesischaracterisedbyequally al.(2010)thatfoundnodifferencesbetweenthesetwohabi- sharpenvironmentalgradients,suchastemperatureandfood tats and partly contradicts the results of Netto et al. (2005) availability.Asanimportantcomponentofcontinentalmar- thatfoundhigherdiversityinbasinssoutheastofBrazilonly gins, slopes are also characterised by high habitat hetero- whennematodegeneradiversitywasconsidered. geneity and host diverse communities (Levin and Sibuet, Microbialbenthiccommunitystudiesarearecentaddition 2012).Incontrast,oceanicbasinslackenvironmentalgradi- to deep-sea research in the Mediterranean. A high level of entsandarerelativelyuniform,appearingsimilartodeserts; bacterial richness has been recorded in the Mediterranean forthatreasontheywerelongconsideredconstantandstable Sea(Lunaetal.,2004;Polymenakouetal.,2005a,b,2009), environments. However, evidence now exists to support the which is comparable to other deep-sea sediments (Li et al., fact that basins are dynamic environments that sustain con- 1999a, b; Bowman and McCuaig, 2003). Contrary to what siderableregularandepisodicdisturbances,suchasseasonal has been found for the larger benthic components, micro- phytodetritus deposition and benthic storms (Rex and Etter, bialcommunityparametersdonotchangewithdepthbutre- 2010). mainconstant(Rexetal.,2006;Danovaroetal.,2010).Re- As most of the deep sea is heterotrophic, food supply cent studies dealing with bacterial diversity estimates have to deep-sea benthos is derived ultimately from surface pro- revealed that bacterial sequence richness found in the deep duction.Primaryproductivityoftheeuphoticzonevariesin seaissimilartoestimatesformicroorganismslivinginsoilor spaceandtimeandsodoestheorganicmatterarrivingatthe shallowwater(Lietal.,1999a,b;Polymenakouetal.,2005a, deepseafloor; thismayin turnlead tospatio-temporalvari- 2009;Kouridakietal.,2010;Schaueretal.,2010;Zingeret ability in the benthos. The variability of benthic communi- al.,2011).Inaddition,thereislittleevidencetosupportubiq- tiesalongthecontinentalmarginsismostlyrelatedtodepth, uitousdispersalorabiogeographicalpatternofsedimentbac- which may, however, reflect changes in food supply, sedi- teria (Polymenakou et al., 2005b). Only recently, Zinger et ment characteristics or other factors. While benthic stand- al.(2011)performedananalysisof9.6millionbacterialV6- ingstocksdecreaseexponentiallydownthecontinentalmar- rRNAampliconsfor509samplesthatspantheglobalocean’s ginandreachextremelylowlevelsintheabyssalplain(Rex surfacetothedeep-seafloorinordertoinvestigateglobalpat- et al., 2006), deep-sea biodiversity is among the highest on ternsofbacterialdiversity.Theiranalysishasshownremark- earth(Ramirez-Llodraetal.,2010andreferencestherein).A ablehorizontalandverticallarge-scalepatternsinmicrobial unimodal diversity-depth pattern with a mid-slope diversity communities. Overall, benthic communities appeared more maximumhasbeenrecognisedforbenthos,thoughthistrend diversethanpelagiccommunities,andasubstantiallyhigher and the depth where diversity peaks are not universal and diversity of bacterial populations was recorded in the deep- mayvaryamongbasins,regions,benthiccomponentsortaxa seasedimentscomparedtoopenoceansurfacewaters,vents (RexandEtter,2010,Ramirez-Llodraetal.,2010,andrefer- andanoxicecosystems(Zingeretal.,2011). encestherein).Hence,increasingthesamplingeffortandex- Toexpandknowledgeofdeep-seacommunityspatialpat- tendingresearchtoincludemoretaxaandmoreareasofthe terns, we examine the dominant small fractions of benthos, deepoceanswillhelptofurtherelucidatebathymetrictrends meiofauna and microbes. A large and detailed dataset from ofbenthicdiversity. thecentral-easternMediterraneanBasinisusedfordescrib- IntheMediterranean,thecontinentalshelfisverynarrow ing and comparing basin and slope ecosystems over large andthereforethelargestpartofthisenclosedseaisclassified scale, including different geographical areas and bathymet- as deep sea. Though Mediterranean ecosystems are among ric ranges. We pay special attention to the potential role of themoststudiedareasoftheworld,deep-seafaunaresearch foodavailabilityfortheobservedpatternsbecauseithasbeen lags behind those of other areas (Danovaro et al., 2010). invoked as a major factor of benthic trends in the Mediter- Nonetheless,considerableworkcarriedoutoverthelastthree ranean (Danovaro et al., 1995; Tselepides et al., 2000; Tse- decadesonthedeepMediterraneanmeiofaunahasadvanced lepides and Lampadariou, 2004; Lampadariou and Tselepi- ourknowledgeonthesmallestbutmostabundantmetazoans des, 2006). More specifically, we aimed at investigating (1) ofthesediments(extensivebibliographicreferencesreported whether meiofaunal, nematode and microbial community in Danovaro et al. (2010) and Gambi et al. (2010)). Fol- characteristics(abundance,diversity,composition)differbe- lowing major trends in ecology, recent investigations have tween basins and slopes; (2) whether there are bathymetric sought for latitudinal, longitudinal and bathymetric patterns patterns for any of these community aspects comparable to ofmeiofauna(LampadariouandTselepides,2006;Danovaro worldwide patterns of deep-sea fauna; (3) whether the ob- et al., 2009a, 2010; Gambi et al., 2010), while few study servedpatternsaresimilarbetweenthetwostudiedhabitats, differences in meiofaunal patterns among deep-sea habitats within different subregions and among the different benthic (Danovaro et al., 2009a, b; Vanreusel et al., 2010; Gambi components; and (4) whether the patterns resulted from the et al., 2010). Danovaro et al. (2009a), in a study including currentsynthesisarerelatedtofoodavailability. Biogeosciences,10,4861–4878,2013 www.biogeosciences.net/10/4861/2013/ K.Sevastouetal.: BenthiccommunitiesinthedeepMediterraneanSea 4863 Table1.Overviewofprojects,surveysandinvestigatedareasinvolvedinthepresentstudy. Project Expedition ResearchVessel Date Area NoofStations Depthrange(m) MATER MATERCruise1 AEGAEO Mar-97 NorthernAegean, 13 115–2273 MATERCruise2 AEGAEO Sep-97 CretanSea MITTELMEER1997/98 METEOR40/3 METEOR Dec-97 NorthernAegean, 8 1221–4261 CretanSea, LibyanSea MATER TransMediterranean AEGAEO Jun-99 CentralMediterranean, 4 2950–3870 EasternLevantine, LibyanSea ADIOS ADIOSCruise2 AEGAEO Oct-01 CentralMediterranean 3 2786–2837 BIODEEP BIODEEPCruise1 AEGAEO Aug-01 CentralMediterranean 5 3080–3424 HERMES HERMES3(HCMR) AEGAEO May-06 CretanSea, 10 508–3603 EasternLevantine, LibyanSea AnaximanderMountains METEOR71/1 METEOR Dec-06 EasternLevantine, 1 1544 LEVAR METEOR71/2 METEOR Jan-07 LibyanSea 3 4138–4392 BIOFUN TRANS-MED SARMIENTODE Jun-09 CentralMediterranean, 5 1204–3335 GAMBOA LibyanSea ZOOTOP MSM14/1 MARIAS.MERIAN Jan-10 EasternLevantine, 13 874–2419 REDECO REDECOCruise1 AEGAEO May-10 CretanSea, 8 1049–3607 REDECOCruise2 AEGAEO May-11 LibyanSea 2 Methods 2.2 Laboratorytreatment 2.1 Studyareaandsampling In the laboratory, meiofauna samples from 69 stations were sievedthrough500/1000and30µmmeshsieves.Thefauna TheMediterraneanSea,anoligotrophicmarinesystemwith retained on the 30µm sieve was extracted by either tripli- increasedsalinity,isthelargestanddeepestenclosedseaon catecentrifugationortriplicateflotation,usinginbothcases earth. The Strait of Sicily divides the Mediterranean into Ludox TM colloidal silica solution of 1.18 specific grav- two basins, the western and the central-eastern basins. It is ity. After extraction, meiofauna samples were stained with characterised by many complex structural features, such as Rose Bengal, counted and sorted to higher taxon level un- canyons,coldseeps,seamountsandmudvolcanoes,andsev- derastereomicroscope.Nematodecommunityanalysiswas eral steep environmental gradients, among which an east- basedonsamplesfrom22outofthe69stations.Fornema- ward increasing salinity and temperature gradient and a tode identifications, subsampling was performed so that at southward and eastward decreasing productivity gradient. least 200 nematodes per sediment core were selected. Af- Due to the narrow continental shelf, the largest part of the ter subsampling, specimens were slowly evaporated in an- Mediterraneanisclassifiedasdeepsea. hydrous glycerol, evenly spread on microscope slides and In the course of 10 multidisciplinary European collabo- identifiedtogenuslevelusingthepictorialkeysofPlattand rations, benthic samples were collected during 12 oceano- Warwick(1983,1988),Warwicketal.(1998)aswellasrel- graphic surveys at different depths in the central-eastern evant literature dealing with new species and genera from basin of the Mediterranean Sea (Table 1). Overall, a grid the Mediterranean (e.g. Schuurmans Stekhoven Jr., 1950; of 73 stations located at slope and basin ecosystems were SoetaertandDecraemer,1989;SoetaertandVincx,1987). sampled at 5 regions along the Mediterranean: the central Total microbial community DNA was extracted from ap- Mediterranean,thenorthernAegeanSea,theCretanSea,the proximately1gofsedimentperstation(UltraCleanSoilkit, LibyanSeaandtheeasternLevantine(Fig.1,TableS1).At MoBio, Carlsbad, CA, USA) and 16S rRNA gene libraries each station, sediment samples were collected by means of weresuccessfullyconstructedfor16ofthesampledstations multiple-core sampler, which allowed for undisturbed sed- (19samples).DNAconcentrationswerequantifiedbyusing iment surface sampling. All analyses were focused on the the NanoDrop ND-1000 UV-Vis Spectrophotometer (Nan- sediment surface, where the bulk of meiofauna metazoan oDrop Technologies, USA). The V5-V6 region of the 16S andmicrofaunaisgathered,i.e.5cmand2cm,respectively. rRNA gene was amplified by PCR. The PCR reaction mix- Meiofaunalsampleswerefirsttreatedwith6%MgCl tore- 2 ture(finalvolumeof15µL)contained5µLof5XKAPAHiFi lax tissues and subsequently preserved by adding 10% for- Fidelity buffer (contains 2.0mM Mg2+ at 1X), 0.75µL of malin.Sedimentsamplesformicrobialandbiogeochemistry KAPA dNTP Mix (10mM each dNTP), ∼10g of template (organic carbon and chloroplastic pigment concentrations) DNAand0.50µLofKAPAHiFiHotStartDNAPolymerase analysiswerekeptfrozenat−20◦C. (1UµL−1) (KAPA Biosystems). The V5-V6 region was www.biogeosciences.net/10/4861/2013/ Biogeosciences,10,4861–4878,2013 4864 K.Sevastouetal.: BenthiccommunitiesinthedeepMediterraneanSea Northern Aegean Cretan Sea Eastern Levantine Central Mediterranean Libyan Sea Fig.1.StudyareasandsamplingstationsalongtheMediterraneanSea. amplified with the following set of primers (covering bac- FollowingWhittaker’sscheme(1972),wecalculatedTRand teria and archaea): 802f (50-GATTAGATACCCBNGTA-30) NRseparatelyforthreelevelsofinventorydiversity(thedi- and 1027r (50-CGACRRCCATGCANCACCT-30) (based on versity of a defined geographic unit): (i) for each station, Claesson et al., 2009). The following thermal cycling pro- identifiedeitherasbasinorslopehabitat,asaunivariatemea- gram was applied: initial denaturation at 95◦C for 5min, sureofalphadiversity;(ii)foreachhabitatwithineachofthe 30 cycles of denaturation at 98◦C for 20s, primer anneal- five studied areas as a univariate measure of gamma (land- ing at 55◦C for 15s, and extension at 72◦C for 30s fol- scape/large area) diversity; and (iii) for the whole data set lowed by a final extension at 72◦C for 5min. Quantifica- thatcorrespondstoepsilondiversity(biogeographicprovince tion of the PCR products was performed using the SYBR diversity).ThefirstorderJacknifeestimator(Jack1)wasalso GreenstainandaQuantiFluorspectrophotometer(Promega). calculatedusingEstimateSv.8.2.0(Colwell,2006)asanes- The sequences of the partial 16S rRNA genes were pro- timatoroftruenematoderichnessforthewholemeiobenthic duced in the laboratories of the Institute of Marine Biol- datasetandforeachhabitat. ogy, Biotechnology and Aquaculture of the Hellenic Centre Differentiationdiversity,i.e.therateofchangeinspecies for Marine Research (Crete, Greece) by using a Roche GS- composition, can be measured in many different ways. Fol- FLX454pyrosequencer(Roche,Mannheim,Germany)fol- lowing the concept of Anderson et al. (2006), we measured lowingtheinstructionsofthemanufacturerforampliconse- beta and delta diversity using multivariate dispersion based quencing.Sequencesthatwereshorterthan200bpinlength on Jaccard dissimilarity. Beta diversity was measured (i) as wereremoved.TaxonomywasassignedusingtheRibosomal the variability in taxa/genera composition among stations Database Project classifier. Pyrosequencing noise was re- per habitat for each depth range/area and (ii) as the vari- movedbyusingthedenoiserprogram.Pyrosequencingdata abilityintaxa/generacompositionamongstationsperdepth were submitted to NCBI Sequence Read Archive with the rangewithineachhabitat.Inasimilarcontext,deltadiversity studyaccessionnumberSRA054862. wasconsideredasthevariabilityintaxa/generacomposition Pigments(chlorophylla(Chla),phaeopigments(Phaeo), amongallstationsforeachhabitatandforeachdepthrange. chlorophyll pigment equivalent (CPE)) were measured fol- Besidesbeingflexibleonthemeasureofdissimilarityused, lowing Lorenzen and Jeffrey (1980). Total organic carbon the method proposed by Anderson et al. (2006) has the ad- (TOC) was measured with a Perkin Elmer CHN 2400 anal- vantageoftestingfordifferencesindifferentiationdiversity yser(HedgesandStern,1984). through a multivariate test for homogeneity in dispersions (PERMDISP)(Anderson,2006).Anotherfacetofdifferenti- 2.3 Dataanalyses ationdiversity,theturnoveroftaxa/generabetweenhabitats andbetweendepthrangeswasalsomeasuredandforconsis- Inthepresentstudy,meiofaunalandnematodediversityrefer tencywaspresentedasJaccarddissimilarity. totaxonrichness(TR,numberofmeiofaunaltaxa)andgenus richness (NR, number of nematode genera), respectively. Biogeosciences,10,4861–4878,2013 www.biogeosciences.net/10/4861/2013/ K.Sevastouetal.: BenthiccommunitiesinthedeepMediterraneanSea 4865 Formicrobialdiversityanalysis,sequenceswereassigned pected, nematodes predominated in all stations with an av- toOTUs(operationaltaxonomicunits)usingtheQIIMEsoft- erage percentage contribution ranging from 72% to 95% wareat3%sequencedivergence(specieslevel)accordingto in slope stations of the Levantine and the Central Mediter- Schloss and Handelsman (2005). Individual-based rarefac- ranean,respectively(Fig.2a).Othermeiofaunataxaofsome tioncurvesatadistancelevelof3%(specieslevel)andthe importancewerecopepods(4–16%),polychaetes(0.2–2%), commonlyusedformicrobialdiversityanalysisChao1rich- tardigrades (0.2–4%) and the group of soft-bodied animals nessestimatorwerecalculatedusingtherarefactioncalcula- (turbellarians,gnathostomulids)(0.5–4%)(Fig.2a). tor(http://www2.biology.ualberta.ca/jbrzusto/rarefact.php). Irrespective of the type of habitat, depth or area, meio- Hypothesis testing for differences in meiofauna abun- faunal abundance followed the pattern of nematode abun- dance, richness (TR and NR) and meiofauna and nematode dance. The lowest values were recorded at basin stations structurewasdoneusingpermutationalmultivariateanalysis in all five areas, while the highest values were observed ofvariance(PERMANOVA).Ofcentralinterestwastherela- at slope stations only in the central Mediterranean and the tionshipofmeiofaunadescriptorswithhabitattype,i.e.basin Libyan Sea (Fig. 3). PERMANOVA results (Table 2) in- and slope. Nevertheless, since it is well known that water dicated differences in metazoan meiofaunal abundance be- depthaffectsmarinesedimentcommunities,itwasnecessary tweenthebasinandslopehabitatsonlyfortwodepthranges toconsiderdepthrangeasafactorintheanalysis.Thus,the (1000–1500mand2500–3000m),butsignificantdifferences overall experimental design consisted of two fixed factors: weredetectedinrelationtodepthandweremorepronounced habitat, with two levels (basin, slope) and depth, with eight for the basin habitat, where meiofaunal abundance changed levels (<500m, 500–1000m, 1000–1500m, 1500–2000m, gradually from the shallower stations (up to 1000m) to the 2000–2500m, 2500–3000m, 3000–3500 and >3500m). p deeperones(>3000m).PERMANOVAanalysisappliedto values were obtained using 9999 permutations of residuals data sets from each of the five regions separately could not underareducedmodel.DISTLMroutine,apermutationalre- distinguishmeiofaunalabundancesbetweenthetwohabitats gressionanalysis(Anderson,2001;McArdleandAnderson, either(statisticsnotshown). 2001; Anderson et al., 2008), was performed for exploring relationships between univariate meiofaunal and microbial 3.2 Diversityandcommunitystructure variables and depth; meiofaunal, nematode and microbial communitystructureanddepth;andunivariatevariablesand 3.2.1 Meiofaunaltaxa communitystructurewithenvironmentalvariablesthatpossi- bledrivebenthicpatterns(forsimplicity,henceforthreferred Overall, 27 meiofaunal taxa were encountered in all stud- to as simple DISTLM, multivariate DISTLM and multiple ied areas and habitats (ε-diversity). Of these, only nema- stepwiseDISTLM,respectively).Thesepermutationalanal- todes and copepods were found at all stations and sam- yseswerebasedonBray–Curtisdissimilaritiesofsquare-root ples, which along with polychaetes, tardigrades and ha- transformed abundances, except from richness, for which lacaroids were found at all areas and habitats. Cnidarians Euclidean distances on untransformed data are considered were found only in slope stations, whereas aplacophorans, more appropriate. To explore whether the observed bathy- cumaceans, echinoderms and scaphopods were restricted to metricandhabitatheterogeneitypatternsapplytothewhole basin stations. The latter were found only at one station Mediterranean,thepermutationalanalyseswerealsoapplied (northernAegean,1224mdepth). to data sets from each area separately. PERMDISP routine Gammadiversityofthefivestudiedregions(Fig.4a)was was applied for measuring and exploring differences in dif- highreachingthehighestvalueintheAegeanSea(24and23 ferentiation diversity (measured as multivariate dispersion). taxaatthenorthernAegeanandtheCretanSea,respectively), Non-metric multi-dimensional scaling (nMDS) ordinations whereas the lowest value was found at the central Mediter- wereusedtoillustratespatialpatternsofcommunitiesstruc- raneanarea(13taxa).WiththeexceptionoftheeasternLev- ture. All analyses were performed with PRIMER v6 with antine,thenumberofmeiofaunaltaxawashigheratthebasin PERMANOVA+add-onsoftware(ClarkeandGorley,2006; habitat.Nevertheless,α-diversityrangewassimilarbetween Andersonetal.,2008). the basin and slope habitat and varied from 2 to 18 in the basinsofthecentralMediterraneanandthenorthernAegean, respectively,andfrom3to16atLibyanslopestations.This 3 Results wassupportedbyPERMANOVAresultsthatshowedstatis- ticallysignificantdifferencesinTRamongdepthrangesbut 3.1 Meiofaunalstandingstocks notbetweenhabitats(Table2). Deltadiversity(Table3)wassimilaratbothhabitats(PER- Averagemeiofaunaldensitiesinthesedimentsurfaceranged MDISP p>0.05) and rather high, suggesting greater vari- from 2±2 ind10cm−2 in the deep basin of the central abilityinmeiofaunaltaxacompositionwithin(>38%)than Mediterranean(2837mdepth)to1249±259ind10cm−2in between habitats (22%). At smaller scale, i.e. within each the basin of the northern Aegean (1271m depth). As ex- depth category, beta diversity of the two habitats (Table 4, www.biogeosciences.net/10/4861/2013/ Biogeosciences,10,4861–4878,2013 4866 K.Sevastouetal.: BenthiccommunitiesinthedeepMediterraneanSea 100% A 95% Others 90% Halacaroidea 85% Rotifera 80% Ostracoda 75% Kinorhyncha 70% Tardigrada 65% Soft bodied 60% Annelida Basin Slope Basin Slope Basin Slope Basin Slope Basin Slope Copepoda Cent Med Cretan Sea Levantine Libyan Sea North Aegean Nematoda 100% B 90% Other 80% Epsilonproteobacteria 70% Betaproteobacteria 60% Deltaproteobacteria 50% Other Proteobacteria 40% Unclassified Archaea 30% 20% Thaumarchaeota 10% Alphaproteobacteria 0% Acidobacteria Slope Basin Slope Basin Slope Gammaproteobacteria Cretan Sea Libyan Sea Levantine Unclassified Bacteria Fig.2.Meiofauna(A)andmicrobial(B)compositionperstudiedareaandhabitat. within habitats) did not differ either (PERMDISP p>0.05 in structure is gradual, as successive depth ranges appear for all depth categories) but appears to increase with depth. similarintermsofmeiofaunataxadownto2000mdepth. Meiofaunaltaxaturnoverbetweenbasinandslope(Table4) increases down to 1500m depth, after which it appears to 3.2.2 Nematodes stabiliseclosetoahighvalue(∼55%). When the different depth ranges were tested (Table 3, A total of 155 genera were encountered in the 22 stations within depth ranges), differences in delta diversity were wherenematodecommunityanalysiswasconducted.Among detected (PERMDISP p<0.01) due to the low variability these, three genera were found in all stations (Acantho- in meiofauna composition among the very shallow stations laimus, Halalaimus, Sphaerolaimus) and six more genera (<500m,13%)(PERMDISPpofpairwisetests<0.05for were found in all areas and habitats (Daptonema, Metas- all comparisons of the first depth range). Statistically sig- phaerolaimus, Microlaimus, Oxystomina, Pselionema, Sy- nificant differences were also found for beta diversity of ringolaimus).Forty-eight(48)generawererestrictedtoone depthrangesforthebasinhabitatbutnotforslope(Table4, station,ofwhich46werefoundonlyinonesample. within depth ranges). Jaccard dissimilarity between depth Similar to TR, γ-diversity of nematodes peaked in the ranges for the whole data set (Table 3) was relatively in- northern Aegean Sea (108 genera), but it was lowest at the creasing for depth categories deeper than 1500m, indicat- most eastern part of the basin, the Levantine (22) (Fig. 4b). ing higher compositional differences in meiofaunal taxa to- Alphadiversityrangedfrom13intheeasternLevantinebasin wards greater depths. This pattern of bathymetric increase habitatto77inthenorthernAegeanslopeenvironmentand, was stronger within the slope habitat but was not observed as opposed to TR, one-way PERMANOVA analysis indi- for basin (Table 4, between depth ranges), for which val- cateddifferencesinthenumberofnematodegenerabetween ues were rather high (89% of values >30%) regardless thebasinandtheslopehabitat(Table2).Thispatterncould ofthedepthcategory. not,however,beverifiedforeachofthestudiedareassepa- Community structure analysis (nMDS and PER- ratelyexceptfromtheCretanSea. MANOVA) confirmed the high variability in meiofauna Thenumberofnematodegenerapredictedbythespecies structure within each habitat and indicated differences richnessestimatorJack1forthewholeinvestigatedarea,the among depth ranges (Fig. 5a, Table 2). Meiofauna compo- basinandtheslopehabitat(mean±SD:160±10,128±12 sition differed between the shallower (up to 1500m depth) and127±11,respectively)isveryclosetotheoneobserved and the deeper stations (>2500m), however, the difference in the data set (155, 123 and 126). The estimator showed a clearsignofapproachinganasymptotefortheslopehabitat only(Fig.6)despitethefactthatthenumberofsamplesfor Biogeosciences,10,4861–4878,2013 www.biogeosciences.net/10/4861/2013/ K.Sevastouetal.: BenthiccommunitiesinthedeepMediterraneanSea 4867 Table2.PERMANOVA(permutationalmultivariateanalysisofvariance)resultsoftheeffectsofhabitatanddepthforuni-andmultivariate meiofaunal variables. p values were obtained using 9999 permutations of residuals under a reduced model. Where statistical significant differenceswereindicatedpairwisecomparisonsweredone.Depthrangeswithcommonunderlinearenotsignificantlydifferentatthe95% significancelevel. Variable Source F p aposterioripairwisecomparisons Meiofaunal Habitat×Depth 2.9307 0.0046 abundance Habitat 0.58247 0.5592 Depth 6.2866 0.0001 Withinhabitat:slope <500 500–1000 1000–1500* 2000–2500 2500–3000* 3000–3500 >3500 1500–2000 <500 500–1000 2000–2500 1000-1500 1500-2000 2500-3000 Majortaxa HabitatxDepth 1.4005 0.2374 richness Habitat 3.1846 0.0768 Depth 7.936 0.0001 <500 500–1000 1000–1500 1500–2000 2000–2500 2500–3000 3000–3500 >3500 Nematode Habitat 5.1425 0.0343 richness1 Meiofaunal HabitatvDepth 1.5549 0.0582 community Habitat 1.3677 0.2243 Depth 4.9251 0.0001 <500 500–1000 1000–1500 1500–2000 2500–3000 2000–2500 >3500 3000–3500 Nematode Habitat 1.6703 0.0983 community1 *statisticallysignificantdifferencesbetweenhabitatsforthespecificdepthrangeatp<0.05. 1One-wayPERMANOVA(depthwasnotincludedintheexperimentaldesign). thishabitatiscomparablylow(12).AsopposedtoNR,Jack1 One-way PERMANOVA analysis based on nematode predicts no difference in nematode genera number between community structure did not indicate differences between basinandslope,astheconfidenceintervalsofthetwocurves basin and slope (Table 2), which is also depicted in the rel- overlap(notshowninFig.6). evant nMDS plot (Fig. 5b). However, nematode communi- Differentiation diversity analysis of nematode commu- ties of most bathyal stations (down to 1000m for slope and nity was performed only at the level of the whole data set 2000m for basin stations) are grouped separately from the (deltadiversity),asatsmallerscales(withineachhabitatand rest, suggesting that depth has an effect on nematode com- within each depth range) not all levels of each factor were munitystructure. present. PERMDISP revealed strong differences in multi- variate dispersions between the two habitats (PERMDISP 3.2.3 Microbes p<0.05) with higher delta diversity recorded at the basin habitat (Table 3, 47%). A rather high Jaccard dissimilarity Pyrosequencing analysis produced 55213 16S rRNA pyro- value was also recorded between the two habitats (39%). tags and a total of 9587 different OTUs were finally iden- In contrast, no differences in delta diversity were detected tified from the 19 samples analysed for microbial diversity. between the different depth ranges (PERMDISP p>0.05), Taxonomicanalysisledtotheclassificationofthesequences which ranged from 18% at the shallowest depth range to in15bacterialand3archaealphylaandto5candidatedivi- 41% at 2500–3000m depth range (Table 3). However, Jac- sions(Fig.2b).Sequenceswerehighlydominatedbyuniden- card dissimilarity between depth categories ranged within tified members of bacteria (34% of total sequences), while high values (41–75%) suggesting high variability in nema- the phylum of Proteobacteria, including the five classes of todegenerabetweendepths. Alpha-,Beta-,Gamma-,Delta-andEpsilon-proteobacteria, wasthemostdominantoftheknownbacterialgroups(32% www.biogeosciences.net/10/4861/2013/ Biogeosciences,10,4861–4878,2013 4868 K.Sevastouetal.: BenthiccommunitiesinthedeepMediterraneanSea Table3.DeltadiversityofmajormeiofaunaltaxaandofnematodegenerabasedonJaccarddissimilarity.Valueswithineachhabitatand withineachdepthrangewerecalculatedusingmultivariatedispersion.Dissimilarityincreasesfrom0to100. Majormeiofaunaltaxa Withinhabitats Basin Slope 42 38 Betweenhabitats 22 Withindepthranges∗∗ <500 500–1000 1000–1500 1500–2000 2000–2500 2500–3000 3000–3500 >3500 13 33 39 39 30 39 36 41 Betweendepthranges <500 500–1000 1000–1500 1500–2000 2000–2500 2500–3000 3000–3500 500–1000 26 1000–1500 31 22 1500–2000 33 17 22 2000–2500 48 36 46 43 2500–3000 29 18 31 18 32 3000–3500 35 39 42 32 39 26 >3500 45 42 38 35 42 38 37 Nematodegenera Withinhabitats∗ Basin Slope 47 41 Betweenhabitats 39 Withindepthranges <500 500–1000 1000–1500 1500–2000 2000–2500 2500–3000 3000–3500 >3500 18 n/a 38 36 28 41 39 30 Betweendepthranges <500 500–1000 1000–1500 1500–2000 2000–2500 2500–3000 3000–3500 500–1000 44 1000–1500 47 47 1500–2000 47 50 48 2000–2500 57 63 56 60 2500–3000 65 66 61 57 53 3000–3500 59 60 56 60 41 51 >3500 75 75 75 72 67 62 61 ∗Statisticallysignificantdifferencesamonghabitatsatp<0.05;∗∗statisticallysignificantdifferencesamonglevelsofthefactor“depthrange”at p<0.001;n/a:FewsamplesforrunningPERMDISP. oftotalsequences).Thenextmostdominantphylumwasthe ranges(>77%)andsamples(>78%)wasveryhigh,clearly Acidobacteria (7% of total sequences) followed by the ar- indicating high variability of microbial communities at all chaealphylumofThaumarchaeota(6%oftotalsequences). spatialscales.ThiswasfurthersupportedbynMDSanalysis Microbial community composition appeared to be similar asthelackofapparentgroupingcouldnotrevealanyspatial between slope and basin stations at the level of large tax- patternorgradient(Fig.5c). onomic groups, except from the slope station in Levantine, where the class of Gammaproteobacteria was highly domi- 3.3 Benthos and environmental variables: bathymetric nant (Fig. 2b). However, this could be a random effect, as trendsandfoodavailability microbialcommunitycompositioninslopesoftheLevantine isbasedontheanalysisofasinglestation. SimpleDISTLManalysiscomplementedPERMANOVAre- Richness was much higher at the basin compared to the sults, indicating that depth explains a significant amount of slopehabitat,accountingfor7969and3329differentOTUs, thevariationinmeiofaunalabundanceandrichness(Table5), respectively, while the Chao1 richness estimator, based on which decrease with increasing water depth (Fig. 7). The thewholedataset,predicted19489differentOTUs(Fig.6). variationexplainedwashigherforbasinascomparedtoslope AsmallproportionofOTUsrangingfrom2to28%weresin- habitat for abundance and TR, but it was not statistically gletons(i.e.occurredonlyonceinthefulldataset)andthey significant for NR when slope stations were only consid- accountedforapproximately10%ofallsequences.Jaccard ered (Table 5). Microbial diversity, on the other hand, was dissimilarity between the two habitats (82%), among depth foundtobeonlymarginallysignificantlydependentondepth whenthewholemicrobialdatasetwasinvestigated(Table5), Biogeosciences,10,4861–4878,2013 www.biogeosciences.net/10/4861/2013/ K.Sevastouetal.: BenthiccommunitiesinthedeepMediterraneanSea 4869 Table4.BetadiversitybasedonmajormeiofaunaltaxaJaccarddissimilarity.Valueswithineachhabitatandwithineachdepthrangewere calculatedusingmultivariatedispersion.Dissimilarityincreasesfrom0to100. β-diversity Withinhabitats Basin Slope 500–1000 23 29 1000–1500 26 37 1500–2000 31 37 2500–3000 40 n/a <500 500–1000 1000–1500 1500–2000 2000–2500 2500–3000 3000–3500 >3500 Betweenhabitats 17 41 50 41 57 56 n/a n/a Withindepthranges <500 500–1000 1000–1500 1500–2000 2000–2500 2500–3000 3000–3500 >3500 ∗ Basin n/a 23 26 31 25 40 36 41 Slope 14 29 37 37 17 n/a n/a n/a Betweendepthranges <500 500–1000 1000–1500 1500–2000 2000–2500 2500–3000 3000–3500 Slope <500 27 36 48 74 63 500–1000 33 19 38 70 60 1000–1500 36 43 22 67 56 n 1500–2000 si 36 36 38 57 53 a B 2000–2500 47 47 54 41 25 2500–3000 26 35 44 23 32 3000–3500 33 42 36 36 39 26 <3500 45 52 39 32 42 38 37 ∗Statisticallysignificantdifferencesamonglevelsofthefactor“depthrange”forthebasinhabitatatp<0.05;n/a:FewsamplesforrunningPERMDISP. suggestingthatmicrobialrichness,oppositetothemeiofau- availabilityusedinthisstudycouldexplainthevariabilityof nalvariables,tendstoincreasewithdepth(Fig.7).However, Mediterraneanmicrobialcommunity. thistrendcouldnotbeverifiedforeachofthetwohabitats. Similar to simple DISTLM, multivariate DISTLM analysis usedtoinvestigatethebathymetricpatternsofbenthiccom- 4 Discussion munitiesindicatedthatdepthexplainsasignificantvariation of the structure of both meiofaunal and nematode commu- 4.1 Standingstocksinthedeepsea nity but not of microbial, although it suggests (p=0.045) Intheenergy-poorenvironmentofthedeepsea,manycom- correlation of microbes with depth when the slope stations munityparametersseemlinkedtofoodsupply.TheMediter- areonlyinvestigated(Table5). ranean Sea has low levels of productivity, particularly the When food availability is also considered, multiple step- Eastern Basin (Psarra et al., 2000), which results in ex- wiseDISTLM(Table6)revealsthatdepthremainsthemost tremelyloworganiccontentindeepsediments(Danovaroet important factor in explaining meiofaunal variability, along al.,1999).Therefore,benthicstandingstocksintheMediter- withoneoftheproxiesoffoodavailabilityforallcasesex- raneanareexpectedtobedepressed.Nevertheless,thereare ceptfromnematoderichness,forwhichdepthaloneexplains areas in the Mediterranean Sea which are more productive. 73% of its variance. Similar results are provided when the Suchanareaisthe northern AegeanSea,whichisenriched basinhabitatisonlyinvestigated,butattheslopehabitatthe by riverine outflows and the influx of nutrient-rich Black importance of food availability is increasing and becomes Sea waters (Poulos et al., 1997; Lykousis et al., 2002). As the sole factor, explaining more than 60% of the variance a consequence, higher faunal densities are anticipated for inmeiofaunalandnematoderichnessandnematodecommu- thisarea.Themeiofaunalabundancesreportedinthepresent nity.Onasimilarnote,depthisalsotheonlyfactorexplain- synthesis are indeed lower from those of similar depths in ing at a high rate (>60%) the variance of microbial rich- other oceans (for an extensive reference report see Gambi ness; yet, this is not verified for the slope habitat, whereas et al., 2010) but comparable to those recorded from simi- sedimentaryorganiccarbonisthesolevariablesuggestedfor larMediterraneandeep-seasediments(TselepidesandLam- explaining microbial richness variability in the basin habi- padariou, 2004; Lampadariou and Tselepides, 2006; Lam- tat.Incontrast,neitherdepthnoranyoftheproxiesoffood padariouetal.,2009;Gambietal.,2010),yettheyfallwithin awiderrangeofvalues.Ourdataalsoconfirmstheprediction www.biogeosciences.net/10/4861/2013/ Biogeosciences,10,4861–4878,2013 4870 K.Sevastouetal.: BenthiccommunitiesinthedeepMediterraneanSea 160 30 Major taxa richness A Central Mediterranean 140 a 25 2 nd/10 cm 1102680000 Number of tax 1250 i 40 10 20 5 0 2786 2799 2837 3080 3200 3281 3301 3334 3335 3424 2011 0 Central Northern Cretan Sea Libyan Sea Eastern 300 Mediterranean Aegean Levantine Cretan Sea 250 120 Nematode richness B 2 m 200 100 c ind/10 11055000 ber of genera 6800 um 40 0 N 914 1194 1284 1580 1772 1839 1840 2273 1049 1619 1898 20 140 0 Eastern Levantine Central Northern Cretan Sea Libyan Sea Eastern 120 Mediterranean Aegean Levantine 100 2 m Overall Basin Slope c 80 0 1 d/ 60 Fig. 4. Meiofauna diversity per studied area and habitat based on n i 40 (A) major meiofaunal taxa, (B) nematode genera. The overall bar 20 indicatestherichnessperareawhenbothhabitatsarecombined. 0 1966 2043 2419 3544 874 942 1465 1499 ofhigherfaunaldensitiesinthenorthernAegeanSea,asthe 450 400 Libyan Sea highermeiofaunaldensitywasmeasuredinthisarea. 350 One of the best known gradients in marine sediments is 2 m 300 that of abundance with water depth. Similar to all benthic d/10 c 220500 groups,therelationshipofmeiofaunaldensityanddepthhas n beenfoundtobesignificantandnegative(Rexetal.,2006). i 150 100 Thisgeneralpatternhasbeenmainlyrelatedtoorganicmat- 50 terinput,foodlimitationandfoodquality(Rexetal.,2006; 0 Soltwedel,2000),althoughitisalsosuggestedtobelocally 2707 2845 3315 3589 3607 4260 508 1079 1204 1910 2980 influencedbyanumberoffactorssuchasthehydrographyof 1600 theareaandthesedimenttype(Gambietal.,2010;Rexand Northern Aegean 1400 Etter, 2010). Our results are in agreement with this general 1200 2 m 1000 bathymetrictrendofabundanceandindicatetheexistenceof 0 c 800 this pattern for meiofauna in all the eastern Mediterranean 1 nd/ 600 basinsirrespectiveofthetypeofhabitat.Nevertheless,itwas i 400 onlyintheslopehabitatthatastrongrelationwithfoodavail- 200 ability was indicated, suggesting that despite both habitats 0 sharingthesamebathymetricpattern,theunderlyingmecha- 115 805 965 1221 1224 1250 1271 153 340 675 nismsdiffer.Itappearsthattheresponseofmeiofaunatothe Depth (m) organicmattercontentinbasinsisaffectedbyotherpossible Basin Slope drivers of meiobenthic patterns, such as food quality, local Fig.3.Nematodeabundanceperstudiedareaandhabitatinrelation hydrodynamics, sediment texture, or even inherent popula- todepth.Noteinx-axisthearrangementofdepthinincreasingorder tioncharacteristics(lifecycle,metabolicactivity,etc.). foreachhabitat. Inapan-Mediterraneandeep-seameiofaunareviewstudy based on primarily literature data from 195 sites, Gambi et al. (2010) reported that meiobenthic abundances at slopeswerehigherthanotherinvestigatedtypesofhabitats, Biogeosciences,10,4861–4878,2013 www.biogeosciences.net/10/4861/2013/

Description:
sity, irrespective of the benthic group or level of taxonomic . 10. 508–3603. Eastern Levantine,. Libyan Sea. Anaximander Mountains METEOR 71/1.
See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.