MARINE ECOLOGY PROGRESS SERIESMar Ecol Prog Ser

Vol. 461: 1–14, 2012doi: 10.3354/meps09894

Published August 8

INTRODUCTION

Certain diatom species are known to release largeamounts of dissolved organic carbon (DOC), mostlyas carbohydrates and other exopolymeric substances(EPS) into the surrounding seawater (Stal 2003).Once released from the cell surface, cell-bound EPS

© Inter-Research 2012 · www.int-res.com*Corresponding author. Email: m.ullrich@jacobs-university.de

FEATURE ARTICLE

Effects of Marinobacter adhaerens HP15 on polymer exudation by Thalassiosira weissflogii

at different N:P ratios

Astrid Gärdes1,4, Yannic Ramaye1, Hans-Peter Grossart2, Uta Passow3, Matthias S. Ullrich1,*

1Jacobs University Bremen, Molecular Life Science Research Center, Campus Ring 1, 28759 Bremen, Germany2Leibniz Institute of Freshwater Ecology and Inland Fisheries, Alte Fischerhuette 2, 16775 Stechlin, Germany

3Marine Science Institute, University of California - Santa Barbara, Santa Barbara, California 93106, USA

4Present address: San Diego State University, Department of Chemistry and Biochemistry, 5500 Campanile Drive, San Diego, California 92182, USA

ABSTRACT: In the ocean, exopolymer accumulationand the resulting aggregation of marine phytoplanktondrive the flux of sinking organic matter. Little is knownabout the contribution of bacteria to the release andbuild-up of such exudates. We describe an in-depth investigation of exopolymer production under differingnutrient conditions using a diatom− bacteria model system. Responses of the marine diatom Thalassiosiraweissflogii to different nutrient re gimes and theimpact of the co-incubated bacterium Marinobacteradhaerens on algal exudation were evaluated by ana-lyzing quantity and quality of exudation products. Cul-tures of T. weissflogii were grown at nutrient-balancedconditions (Redfield ratio N:P = 16) as well as at nitro-gen- or phosphorus-depleted conditions (N:P = 1 andN:P = 95, respectively) in the presence or absence ofM. adhaerens. The impact of M. adhaerens on the con-centration of dissolved organic carbon and transparentexopolymer particles (TEP), as well as on the composi-tion of amino acids and carbohydrates, depended onthe nutrient regime. Under nutrient-balanced con -ditions, M. adhaerens stimulated both T. weissflogiigrowth and TEP production. Under nutrient-depletedconditions, TEP production was enhanced regardlessof whether bacteria were present. Phosphorus limita-tion resulted in an appreciable change in the %molcomposition of dissolved amino acids in xenic and axenic treatments. Differential lec tin staining revealedthat the presence of M. adhae rens enhanced and modified the pro duction of specific extracellular sub-stances. A better understan ding of the effects of diatom−bacteria interactions on the accumulation ofexudation products is crucial in modelling and predict-ing consequences of environmental changes on oceanicorganic matter and nutrient cycling.

Scanning electron micrograph of a co-incubation sample con-taining the diatom Thalassiosira weissflogii (left) and Marino -bacter adhaerens HP15 embedded in a network of chitin- containing fibres extruding from the diatom cell.

Image: A. Gärdes

OPENPEN ACCESSCCESS

KEY WORDS: Diatom−bacteria interactions · Transparentexopolymer particles · TEP · Dissolved organic carbon ·Nutrient limitation · Lectin staining · Marino bacter adhae -rens · Thalassiosira weissflogii

Resale or republication not permitted without written consent of the publisher

Mar Ecol Prog Ser 461: 1–14, 2012

and EPS fibrils (Laspidou & Rittmann 2002) can formnano-gels (Chin et al. 1998) and transparent exopoly-mer particles (TEP) (Passow & Alldredge 1994a, Pas-sow 2000). TEP, which are rich in acidic polysaccha-rides and are retained as Alcian Blue-stainableparticles on filters (Alldredge et al. 1993), promoteaggregation by providing the matrix or ‘glue’ ofmicro- and macro-aggregates (Alldredge et al. 1993,Logan et al. 1995, Passow 2002a). Aggregates con-tribute significantly to the sinking flux of carbon outof the euphotic zone (Fowler & Knauer 1986). It isthus essential to understand conditions that lead tothe accumulation of EPS.

Production of exudates by marine phytoplankton isknown to be highly variable depending on species,growth stage, and environmental conditions (Myk-lestad 1974, 1977). Nitrogen depletion, and for somespecies, phosphorus depletion, can result in an in -creased extra-cellular release of photosynthesis pro -ducts by cultured diatoms (Obernosterer & Herndl1995, Magaletti et al. 2004, Pierre et al. 2012).

Heterotrophic bacteria interact with phytoplanktonin the so-called ‘phycosphere,’ which is the micro-zone surrounding algal cells (Bell & Mitchell 1972).The type of interaction, e.g. competition or mutualism,often depends on nutrient availability (Grossart 1999,Lovdal et al. 2008). Thus, depending on the type of interaction, diatom-associated bacteria may havethe potential to impact the release of algal exudates(Bruckner et al. 2011), TEP formation (Bhaskar et al.2005, Grossart et al. 2006), aggregation (Gärdes et al.2011), and sedimentation (Smith et al. 1995). On theone hand, heterotrophic bacteria cause hydrolysis ofalgal products (Bidle & Azam 1999, Grossart et al.2007, Sapp et al. 2008) resulting in reduced EPS accu-mulation. On the other hand, EPS including TEP arecolonized, formed, utilized, and modified by bacteria(Passow & Alldredge 1994b, Mari & Kiørboe 1996,Verdugo et al. 2004), providing bacterial niches aswell as substrates (Azam & Malfatti 2007). Phyto-plankton exudation may have evolved as a mecha-nism to establish mutualism between heterotrophicbacteria and phytoplankton, in particular by attractingbacteria as EPS re-mineralizers (Azam & Cho 1987).

The role that bacteria−diatom interactions playwith regards to the production of exopolymers andthe impact of nutrient availability in determiningthese interactions are largely unknown. Here, wepresent a series of in vitro experiments using themarine diatom Thalassiosira weissflogii and the bacterium Marino bacter adhaerens HP15 (hereafterMarino bacter adhaerens) as a model to investigatenutrient-dependent exopolymer production by dia -

toms in the presence or absence of diatom-associatedbacteria. M. adhaerens, which was isolated frommarine aggregates (Grossart et al. 2004), preferen-tially attaches to the surface of T. weissflogii andspecifically contributes to TEP formation and algal aggregation (Gärdes et al. 2011). The bacterium hasbeen taxonomically characterized (Kaeppel et al.2012); its genome sequence has been fully annotated(Gärdes et al. 2010), and its genetic accessibility hasbeen demonstrated (Sonnenschein et al. 2011), ren-dering it a suitable model organism. To evaluate theresponse of T. weissflogii to different nutrient condi-tions and the concurrent impact of M. adhaerens onpolymer exudation, the dynamics of DOC and TEP,as well as the composition and amount of releasedpolysaccharides and amino acids, were measured.We hypothesized that M. adhaerens affects quantityand quality of the released algal exudates in a nutri-ent-dependent manner.

MATERIALS AND METHODS

Diatom batch cultures, bacteria, and experimentalsetup

Axenic Thalassiosira weissflogii (CCMP 1336) wasobtained from the Provasoli-Guillard National Cen-ter for Culture of Marine Phytoplankton (CCMP,Maine, USA) and routinely grown in sterile auto-claved pre-filtered f/2 medium (Guillard 1975) basedon artificial seawater (ASW) in 150 ml cell cultureflasks. The cultures were checked regularly for bac-terial contamination by microscopic observation andagar plating on marine broth (MB; ZoBell 1941). Dia -tom cells were quantified by cell counting as de -scribed below.

Five liters of axenic diatom cultures were grownat 16°C in f/2 medium using a 12 h light period at150 µmol photons m−2 s−1 for 9 d. Thalassiosira weiss-flogii cells were harvested by centrifugation (4139 × g,10 min) and subsequently washed with sterile ASWto minimize nutrient carry-over. Aliquots of approxi-mately 5000 T. weissflogii cells were used to inocu-late 6 of 7 experimental treatments (600 ml mediumin 1 l cell culture flasks). The control treatment didnot contain diatoms. The experimental setup is out-lined in Fig. 1. Macro-nutrients in the medium wereadjusted as follows. In treatments I and II, whichwere designated as ‘nutrient-balanced,’ nutrientswere adjusted to 580 µM silicate, 425 µM nitrogen,and 26 µM phosphorus (N:P = 16). In the medium fortreatments III and IV, nitrogen (26 µM) was reduced

2

Gärdes et al.: Bacterial effects on diatom polymer exudation

to an N:P ratio of 1 and termed ‘nitrogen-depleted’(Table 1). In contrast, treatments V and VI containedmedium depleted in phosphate (4.5 µM) with an N:Pratio of 95 and hence was designated as ‘phosphorus-depleted’ (Table 1). Treatments I, III, and V werekept axenic, whereas the bacterial strain Marinobac-ter adhaerens was added to treatments II, IV, and VI(xenic treatments). Control treatment VII containednutrient-balanced medium without diatoms but withM. adhaerens. The bacterium was routinely grownin MB (ZoBell 1941) in 100 ml Erlenmeyer flasksshaken at 150 rpm at 18°C. To avoid carry-over of MBand bac terial metabolites, M. adhaerens cells werewashed and incubated overnight in ASW (Grossart1999) at 18°C. Thereafter, 1 ml of the bacterial sus-pension was added to the respective treatments togive a final density of ~3 × 106 cells ml−1.

All treatments were incubated at 16°C using a 12 hlight period at 150 µmol photons m−2 s−1 for 7 d. Sub-samples of 20 ml were taken with sterile syringes onDays 0, 4, and 7 and subjected to microbial and bio-chemical analyses.

Determination of diatom and bacterial cell concentrations

Cell concentrations of diatoms were determined bycell counts in a Sedgwick-Rafter Cell S50 (SPI Sup-plies) using an inverted Axiovert 200 microscope(Zeiss). Bacterial cell numbers in the initial inocula

were determined by serial dilution plating andthroughout the experiment by microscopic enumera-tions of 4’,6-diamino-2-phenylindole (DAPI)-stainedbacteria. In preparation for the DAPI counts, 200 µlsamples were diluted 1:2 and 1:10 in sterile f/2medium, fixed with 0.2 µm pore size pre-filtered,borate-buffered 2% formalin, and stained with DAPI(1 µg ml−1) for 3 min (Porter & Feig 1980). Sampleswere filtered onto 0.2 µm pore size polycarbonateNucleopore filters (Whatman) and stored at −20°C.Bacterial cell numbers were counted in replicatesusing an epifluorescence Axioplan microscope (Zeiss)at 1000× magnification.

Nutrient analysis

Concentrations of NO3− and PO4− were determinedspectrophotometrically at all 3 sampling points usingan Auto Analyzer Evolution III apparatus (AllianceInstruments). Ten ml samples were pre-filtered (0.2µm Nucleopore) and kept at −20°C until analysis.Determination of nutrients was performed in dupli-cates according to an established method (Strickland& Parsons 1972).

Dissolved organic carbon (DOC)

For DOC analysis, replicates of 10 ml sampleswere collected in pre-combusted glass ampoules

3

Fig. 1. Experimental set-up. Aliquots of an axenic Thalassiosira weissflogii culture weretransferred to nutrient-balanced (N:P = 16) nitrogen-depleted (N:P = 1), and phosphorus-depleted (N:P = 95) treatments (see ‘Materialsand methods’ for concentration) and incu-bated without (axenic) or with the bacteriumMarinobacter adhaerens HP15 (xenic). As acontrol, M. adhaerens HP15 without diatomswas incubated in nutrient-balanced condi-tions. Two replicates were used per treatmentor control. DOC: dissolved organic carbon;DFAA: dissolved free amino acids; DCAA:dissolved combined amino acids; DFCHO:dissolved free neutral and acidic monosac-charides; DCCHO: dissolved combined neu-tral and acidic polysaccharides; TEP: trans-

parent exopolymer particles

Mar Ecol Prog Ser 461: 1–14, 2012

(Carl Roth) after filtration through pre-rinsed 0.2 µmpolycarbonate membranes (Nucleopore). Sampleswere acidified with 1% phosphoric acid, flame-sealed, and stored at −20°C (Grossart et al. 2006).DOC analysis was performed using a high-temper-ature catalytic combustion TOC-5000 instrument(Shimadzu). Given DOC concentrations are aver-ages of 3 injections from each sample and samplingpoint.

Amino acids

Ten ml samples from each sampling point (Days 0,4, and 7), were filtered through 0.2 µm pore size low-protein binding Acrodisc syringe 0.2 µm filters withHT Tuffryn® membrane (Sigma-Aldrich) and storedat −20°C until analysis. Concentrations of dissolvedfree amino acids (DFAA) were analyzed by HPLCafter pre-column derivatization with ortho-phthal-dialdehyde (Mopper et al. 1992). Dissolved combinedamino acids (DCAA) were hydrolyzed with 6 N HClat 155°C for 1 h and analyzed as DFAA (Rosenstock &Simon 1993).

Carbohydrates

Ten ml samples from each sampling point (Days 0,4, and 7) were filtered through 0.2 µm pore size poly-carbonate filters (Nucleopore) and stored at −20°C in15 ml sterile falcon tubes. Dissolved free neutral andacidic monosaccharides (DFCHO) were analyzed byHPLC (Dionex) with a Carbopac PA 10 column(Dionex) and pulsed amperometric detection, using agradient of 3 eluents (10 mM NaOH, 250 mM NaOH,and 25 mM NaOH + 1M Na-acetate buffer) accord-ing to the manufacturer’s protocol. Dissolved com-bined neutral and acidic polysaccharides (DCCHO)were analyzed by HPLC as DFCHO after 20 h ofhydrolysis with 0.09 N HCl at 100°C (Mopper et al.1992).

Transparent exopolymer particles (TEP)

For quantitative TEP analysis, replicate samples of2 to 5 ml from each sampling point (Days 0, 4, and 7)were filtered onto 0.2 µm pore size polycarbonatemembranes, which were subsequently stained withAlcian blue (8GX, final concentration 0.05% in Na-acetate buffer at pH 5.8; Alldredge et al. 1993). Fil-ters were dried and placed on Cytoclear TM slides(Poretics). Twenty randomly selected grids wereexamined by light microscopy at 200× magnification.Image Analysis V 3.0 software (Soft Imaging System)was used to enumerate TEP in size classes, determin-ing total area and size distribution. In total, 300 to700 TEP per slide were measured. The majority ofTEP observed were small, ovate particles. TEP weregrouped into geometric size classes based on maxi-mal length. The maximal and minimum length axeswere used to calculate the elliptical area of TEP ineach size class. The size spectrum from 20 grids andthe average size in each size class were used to cal-culate the total area of TEP, which is a proxy for bio-mass (Passow 2002a). TEP size frequency distributionwas calculated after 7 d.

Lectin staining

Two lectins were used to test for their binding tospecific diatom-derived extracellular polymers: con-canavalin A-fluorescein isothiocyanate (FITC-ConA;Sigma), specific for glucose and mannose residues,and Bandeiraea simplicifolia lectin-tetramethyl rho-damine isothiocyanate (TRITC-Bandeiraea, Sigma),specific for galactose and N-acetylgalactose residues.

4

Treatment Incubation NO3− PO43−

day (µm) (µm)

Balanced nutrient conditions (N:P = 16) Axenic 1 424 ± 26 25.1 ± 6

4 358 ± 11 8.35 ± 2.57 341 ± 16 4.35 ± 2.7

Xenic 1 439 ± 41 26.1 ± 7.64 385 ± 9 7.5 ± 27 366 ± 37 3.9 ± 1.1

Nitrogen-depleted conditions (N:P = 1) Axenic 1 26.3 ± 1.8 29 ± 8.5

4 0.55 ± 0.06 23 ± 67 0.7 ± 0.1 23 ± 9.9

Xenic 1 24.8 ± 2.2 27 ± 4.54 0.42 ± 0.09 25 ± 5.87 0.69 ± 0.1 24 ± 6.1

Phosphorus-depleted conditions (N:P = 95) Axenic 1 464 ± 22 4.8 ± 0.9

4 414.8 ± 27 0.51 ± 0.17 404.8 ± 15 0.18 ± 0.1

Xenic 1 410.4 ± 10 5.57 ± 14 368 ± 25 2.25 ± 0.67 370 ± 31 1.98 ± 0.9

N:P = 16 Control 1 466 ± 37 26 ± 4.24 425 ± 21 15.5 ± 1.27 370 ± 26 10.9 ± 3.3

Table 1. Concentrations of NO3− and PO43− in the differenttreatments and control after 1, 4, and 7 d. Data are mean ±

SD of replicates

Gärdes et al.: Bacterial effects on diatom polymer exudation

Both lectins were applied according to Liener et al.(1986) with modifications as published by Wiggles -worth-Cooksey & Cooksey (2005) at the initial(Day 0) and final sample points (Day 7) in sterile8-well cover slip chambers (Ibidi). Two hundred µlper sample were stained with lectin concentrations of0.1 mg ml−1 in f/4 medium and incubated for 1 to 2 hin darkness. Unbound lectins were removed byadding 100 µl of bovine serum albumin (BSA; 1% inphosphate-buffered saline buffer). In all unstainedcontrols, lectin solutions were replaced by BSA.FITC-ConA and TRITC-Bandeiraea binding activi-ties were determined by confocal laser scanningmicroscopy (CLSM) using an Axiovert 400 (Zeiss)with 1 argon and 1 krypton laser (excitation wavelengths of 488 and 568 nm, respectively, and emis-sion wave lengths of 350 and 560 nm, respectively).Autofluorescence of chlorophyll a was detected usingan excitation wave length of 488 nm and an emissionwave length of 655 nm. Images were processed usingthe Zeiss LSM Image Browser.

Statistical analyses

Simple regression analysis was carried out to assessthe relationships between measured parameters.One-way analysis of variance (ANOVA) was used toassess the differences between DOC, DFAA, DCAA,DFCHO, DCCHO, and TEP, as well as algal and bac-terial cell numbers in xenic and axenic treatments andbetween balanced and depleted treatments.

RESULTS

Nutrient concentrations

At the end of the experiment, after 7 d, nitrogenconcentrations were lower under axenic nutrient-bal-anced conditions as compared to the respective xenictreatment, indicating either a decreased nitrogen up-take by Thalassiosira weissflogii or an additional sup-ply of nitrogen by bacterial remineralization (Table 1).At this time, nitrogen concentrations were low in bothaxenic and xenic treatments under nitrogen-depletedconditions. Under phosphorus-depleted conditions,axenic treatments had used up all available phospho-rus, whereas in xenic treatments, final phosphorusconcentrations were 2 µM (Table 1). In the controltreatment, nitrogen concentration decreased almostas much as in the diatom-containing treatment, butphosphorus concentrations decreased less.

Phytoplankton and bacterial abundances

Thalassiosira weissflogii cell numbers increasedexponentially over the entire 7 d of incubation inboth axenic and xenic nutrient-balanced treatments(Fig. 2A). The increase in algal cell numbers waslargest in xenic cultures between Days 4 and 7(from 1.6 × 104 ± 3.2 × 102 [SD] cells ml−1 to 4.7 ×104 ± 6.8 × 102 cells ml−1). Moreover, in nutrient-balanced treatments, final diatom numbers were sig-nificantly higher (p < 0.05, n = 6) when Marinobacteradhaerens cells were present. In these treatments,bacterial cell numbers increased rapidly, reaching8.8 ± 1.1 × 106 cells ml−1 within 4 d and remainingnearly constant thereafter (Fig. 3).

In contrast, in nitrogen- as well as in phosphorus-depleted conditions (N:P = 1 and N:P = 95, respectively),diatom growth was significantly lower (p < 0.05, n = 6)than in nutrient-balanced treatments. Diatom cellnumbers increased from 5 × 103 ± 0.42 × 103 cellsml−1 to only 2.1 × 104 ± 3.3 × 103 cells ml−1 unaffectedby the presence of Marinobacter adhaerens under nutrient deprivation (Fig. 2B,C). In nitrogen- andphosphorus-depleted treatments, M. ad haerens cellnumbers reached a final concentration of 6.3 ± 0.9 ×106 cells ml−1 and 5.9 ± 2.1 × 106 cells ml−1, respectively(Fig. 3), both significantly lower than under nutrient-balanced conditions (p < 0.05, n = 6).

In the bacterial control treatments, however, bacte-rial abundances decreased by 2 orders of magnitudefrom initially 3.3 ± 1.0 × 106 cells ml−1 to 3.8 ± 0.8 × 104

cells ml−1 at the end of the incubation, indicating thatbacteria lacked a suitable carbon source such asalgal exudates.

Dynamics of dissolved organic matter

In nutrient-balanced treatments, DOC concentra-tions differed significantly (p < 0.05) be tween axenicand xenic treatments (Fig. 4A). Higher DOC concen-trations occurred when diatom cells grew withoutbacteria (axenic): initial DOC con centrations in -creased from 150 ± 17.1 µM C to 600 ± 43.2 µM Con the fourth day of the incubation and remainedhigh thereafter. In contrast, DOC concentrations in -creased to only 400 ± 12.8 µM C in xenic nutrient-balanced treatments (Fig. 4A), suggesting that bac -teria took up substantial amounts of DOC. In thenitrogen-depleted treatment, however, no differencesin DOC concentrations between axenic and xeniccultures were observed, and DOC concentrationsdeclined to 107 ± 20.9 µM C irrespective of the

5

Mar Ecol Prog Ser 461: 1–14, 2012

presence of bacteria (Fig. 4B). In the phosphorus-depleted treatment, high DOC concentrations weremeasured (average = 470.5 ± 31.2 µM C) regardlessof bacterial presence, and DOC remained constantover time (Fig. 4C). The 1-way ANOVA revealedthat DOC concentration over time differed signifi-cantly only between axenic and xenic treatmentsunder nutrient-balanced conditions, and showeda significant difference between nutrient-balancedand depleted conditions only in the xenic treatments(Table 2). This suggested that bacterial DOC utiliza-tion de pended on nutrient availability and was high-est under nutrient-balanced conditions.

Concentrations of DFCHO were below 0.25 µmoll−1 at all sampling points and did not differ among thedifferent nutrient treatments or axenic and xenic cul-tures. Hydrolysis and HPLC analyses revealed glu-cose as the most abundant monomer in all exudates(data not shown).

DCCHO concentrations were below 0.1 µmol l−1 inboth nutrient-balanced and phosphorus-depletedtreatments, both in the presence or absence of bacte-ria. In nitrogen-depleted treatments, DCCHO were3.8 and 2.1 µmol l−1 in axenic and xenic cultures,with glucose accounting for 65 and 45% (2.5 and0.95 µmol l−1), respectively. Other monosaccharides inthe DCCHO pool included fucose, ribose, rhamnose,arabinose, galactose, and mannose (data not shown).The 1-way ANOVA revealed no significant differ-ences in final DFCHO concentrations between mosttreatments, but DCCHO significantly (p < 0.05) dif-

6

Fig. 2. Total proxy of abundance of transparent exopolymerparticles (TEP; filled symbols) and cell density of Thalassio -sira weissflogii (empty symbols) grown under (A) nutrient-balanced (N:P = 16), (B) nitrogen-depleted (N:P = 1), and (C)phosphorus-depleted treatments (N:P = 95) without (axen ic;squares) or with Marinobacter adhaerens HP15 (xenic; circles). Values are means between replicates with SD as error bars. SDs are smaller than the symbol in some cases

Fig. 3. Marinobacter adhaerens HP15. Cell density (mean ±SD of 3 replicates) in the 3 xenic treatments, as well as in the

bacteria-only control

Gärdes et al.: Bacterial effects on diatom polymer exudation

fered between axenic and xenic cultures in nitrogen-depleted treatments (Table 2). This suggested thatpresence of M. adhaerens may lead to lower DCCHOconcentrations when nitrogen availability is low.

Concentrations of DFAA and DCAA were mea-sured in 1 of the duplicate treatments. DFAA concen-trations in the nutrient-balanced treatmentsincreased in both axenic and xenic cultures, from 0.4to 0.62 µmol l−1 and from 0.1 to 0.68 µmol l−1, respec-tively (Fig. 5A). In nitrogen-depleted treatments,DFAA concentrations on Day 7 were substantiallylower (≤0.06 µmol l−1) in the axenic treatment thanin the xenic one. In the xenic treatment, DFAAfirst increased to 0.87 µmol l−1 on Day 4, and thendeclined to 0.44 µmol l−1 at the end of the experiment(Fig. 5B). Maximum DFAA concentrations wereobserved in the phosphorus-depleted xenic (1.9 µmoll−1) and axenic (1.4 µmol l−1) treatments (Fig. 5C). The1-way ANOVA revealed that in nitrogen- and phos-phorus-depleted treatments, significantly higherDFAA concentrations were measured compared tonutrient-balanced conditions in both axenic (p <0.001) and xenic (p < 0.001) treatments (Table 2), butit showed no difference between xenic and axenictreatments.

The average mol-composition (mol%) of DFAA inall treatments, irrespective of inorganic nutrients orbacterial presence, remained similar over time. As -partate, glutamate, tyrosine, and serine represented8–2.5, 10–14.5, 4–10, and 38–45% of DFAA, respec-tively (data not shown).

Under nutrient-balanced conditions, DCAA inaxenic cultures increased and reached concentra-

tions of up to 3.3 µmol l−1. In xeniccultures, higher DCAA concen-trations were observed, with peakvalues of 4.2 µmol l−1 on Day 4(Fig. 5A). In axenic, nitrogen-depleted treatments, DCAA con-centrations peaked on Day 4(3.6 µmol l−1). In the correspond-ing xenic treatment, DCAA de -creased moderately from 3.5to 3 µmol l−1 (Fig. 5B). In axenic,phosphorus-depleted treatments,DCAA increased to 3.9 µmol l−1.In contrast, nearly constant con-centrations of ca. 2 µmol l−1 wereobserved in the xenic, phospho-rus-depleted treatments (Fig. 5C).DCAA concentrations differedsignificantly between xenic nu -trient-balanced and nutrient-

depleted treatments (p < 0.05; Table 2). Axenic treat-ments differed significantly from xenic treatmentswhen nutrients were balanced (p < 0.05; Table 2).

Final DCAA composition in both axenic and xenictreatments under nutrient-balanced or nitrogen-depleted conditions were characterized by highestmol% of alanine (22 to 29% and 33 to 27%, respec-tively) and glycine (12 to 16% and 12 to 15%, respec-tively; Fig. 6). In phosphorus-depleted treatments,phenylalanine and leucine comprised the highestmol% values in both axenic and xenic treatments (23to 27% for phenylalanine and 23 and 15% forleucine). Additionally, phosphorus-depleted treat-ments differed significantly from nutrient-balancedand nitrogen-depleted treatments, respectively, whenaxenic (p < 0.05) and xenic (p < 0.05) treatments werecompared (Table 2). This suggested that changes inquantity and quality of accumulated DCAA weretriggered by nutrient availability.

TEP formation

In nutrient-balanced treatments, final TEP concen-trations differed significantly between axenic andxenic treatments (p < 0.05; Table 2). No TEP formationwas observed in axenic treatments (2.1 ± 0.2 × 109 µm2

ml−1), but TEP increased to 8.0 ± 1.2 × 109 µm2 ml−1 inxenic treatments (Fig. 2A). There was no significantdifference in TEP accumulation rate between axenicand xenic treatments in any of the nutrient-depletedtreatments (Table 2), and the average TEP concentra-tion was 7.5 ± 1.9 × 109 µm2 ml−1 (Fig. 2B,C).

7

Axenic vs. xenic Balanced vs. treatments nutrient-depleted

N:P = 16 N:P = 95 N:P = 1 Axenic Xenic

DOC (µmol C l−1) 0.024 NS NS NS 0.032DFAA (µmol l−1) NS NS NS 0.007 0.001DCAA (µmol l−1) 0.033 NS NS NS 0.05DFAA %mol composition NS NS NS NS NSDCAA %mol composition NS NS 0.0009 0.039 0.042DFCHO (µmol l−1) NS NS NSDCCHO (µmol l−1) NS 0.033 NS NS NSTEP (µm2 ml−1) 0.046 NS NS 0.089 NSTEP size frequency distribution 0.044 NS NS 0.009 NSDiatoms (cells ml−1) 0.01 NS NS NS 0.036Bacteria (cells ml−1) – – – – 0.048

Table 2. p-values of 1-way ANOVA testing for significant differences between ax-enic and xenic treatments and between nutrient regimes on Day 7. DOC: dissolvedorganic carbon; DFAA: dissolved free amino acids; DCAA: dissolved combinedamino acids; DFCHO: dissolved free neutral and acidic monosaccharides; DCCHO:dissolved combined neutral and acidic polysaccharides; TEP: transparent polymer

particles; NS: not significant; (–) not applicable

Mar Ecol Prog Ser 461: 1–14, 2012

After 7 d under nutrient-balanced conditions, sig-nificantly (p < 0.05) more TEP occurred in the sizeclass of 200 to 1000 µm2 in the xenic than in theaxenic treatments, a large fraction of TEP < 200 µm2

being observed in the latter class. The proportion ofTEP >1000 µm2 was highest in the xenic, nutrient-balanced treatment, contributing almost 8% of totalTEP numbers (Fig. 7). Under nutrient-depleted con-ditions, the difference between xenic and axenictreat ments was less pronounced. All nutrient-de -pleted treatments revealed major proportions of TEPin the size class between 200 and 1000 µm2 (p < 0.05),with size frequency distributions similar to those inthe xenic nutrient-balanced treatments. Calculationof TEP production per diatom cell revealed high cell-specific TEP concentrations in xenic, phosphorus-

depleted treatments, with similarly high values in the2 axenic nutrient-depleted treatments and more than2 orders of magnitude lower values in the xenic,nitrogen-depleted treatment (Table 3).

8

Fig. 4. Concentrations of dissolved organic carbon (DOC) in(A) nutrient-balanced (N:P = 16), (B) nitrogen-depleted (N:P= 1), or (C) phosphorus-depleted treatments (N:P = 95) with-out (axenic) or with Marinobacter adhaerens HP15 (xenic).DOC in the bacteria-only control is depicted in panel A

(triangles)

Fig. 5. Concentrations of dissolved free amino acids (DFAA)and dissolved combined amino acids (DCAA) in (A) nutri-ent-balanced (N:P = 16), (B) nitrogen-depleted (N:P = 1), or(C) phosphorus-depleted treatments (N:P = 95) without

(axenic) or with Marinobacter adhaerens HP15 (xenic)

Gärdes et al.: Bacterial effects on diatom polymer exudation

Lectin-stainable exopolymers

To gain better knowledge of the origin and nature ofproduced exopolymers, fluorescence-assisted stainingof samples with FITC-ConA and TRITC-Bandeiraeawas conducted, and lectin staining was visualized by

CLSM. After 1 d of incubation (Fig. 8), ConA (green)was observed as a thin layer on diatom surfaces andalso bound to free particles in the axenic, nitrogen-depleted treatment only (Fig. 8C). Staining withTRITC-Bandeiraea did not yield signals in any of thetreatments on Day 1 of the incubations.

After 7 d of incubation (Fig. 8), FITC-ConA-stain-ing on diatom surfaces was visible in xenic, nutrient-balanced, and nitrogen-depleted treatments (Fig. 8H,J),as well as in both xenic and axenic phosphorus-depleted treatments (Fig. 8K,L) In contrast, Thalas-siosira weissflogii cells did not reveal any FITC-ConA-stained cell surface bound signals in axenic,nutrient-balanced (Fig. 8G) and axenic, nitrogen-depleted (Fig. 8J) treatments, indicating the absenceof polymeric material with glycosidic residues. FreeFITC-ConA-stainable material was found abun-dantly in both axenic and xenic, nitrogen-depletedtreatments (Fig. 8I,J) as well as in the xenic nutrient-balanced treatment (Fig. 8H), but not in the othertreatments. A strong lectin TRITC-Bandeiraea (yel-low) signal overlapped with the FITC-ConA signal inthe xenic, nutrient-balanced treatment (Fig. 8H), butwas not seen in any of the other treatments. Theappearance of TRITC-Bandeiraea in association withFITC-ConA suggests that the FITC-ConA-stainedexopolymer had been either modified by bacterialactivity or a secondary exopolymer had been releasedby the diatom in the presence of bacteria under nutri-ent-balanced conditions.

DISCUSSION

In the 6 treatments, the diatoms were grown under3 different nutrient regimes in the presence orabsence of a bacterial strain known to be associatedwith this diatom (Gärdes et a. 2011). The nutrientregimes differed in their N:P stoichiometry (Redfieldstoichiometry versus nitrogen or phosphorus deple-tion), but also in effective nutrient availability duringthe experiment. Nitrogen and phosphorus remainedavailable throughout the experiment in the 2 nutri-ent-balanced treatments, whereas nitrogen concen-

9

Fig. 6. Mol% composition of dissolved combined aminoacids (DCAA) from the 6 treatments after 7 d of incubation.Amino acids are ASP: aspartate; GLU: glutamate; HIS: histi-dine; SER: serine; GLY: glycine; THRE: threonine; ALA: ala-nine; TYR: tyrosine; VAL: valine; PHE: phenylalanine; ILE:

isoleucine; LEU: leucine

Fig. 7. Relative size distributions of transparent exopolymerparticles (TEP) in the 6 treatments after 7 d of incubation par-titioned into 3 size classes: 10−200 µm2, 200−1000 µm2, and

>1000 µm2

TEP per diatom cell (104 µm2 cell−1)N:P = 16 N:P = 1 N:P = 95

Axenic 8.0 ± 0.14 28 ± 0.09 26 ± 0.28Xenic 16 ± 0.34 0.053 ± 0.60 38 ± 0.51

Table 3. Diatom normalized transparent polymer particle(TEP) concentration in the 6 treatments. Date are mean ± SD

Mar Ecol Prog Ser 461: 1–14, 201210

Fig. 8. Staining with FITC-ConA (green) and TRITC-Bandeira (yellow) after (A–F) 1 d and (G–L) 7d of incubation visualizedwith confocal laser scanning microscopy in (A,B,G,H) nutrient-balanced (N:P = 16); (C,D,I,J) nitrogen-depleted (N:P = 1) (C)axenic and (D) xenic; and (E,F,K,L) phosphorus-depleted (N:P = 95) treatments. Red auto-fluorescence of chlorophyll is also

visible. Arrows: examples of diatom cells, free exopolymers (EPS), and cell-surface-bound EPS

Gärdes et al.: Bacterial effects on diatom polymer exudation

tration dropped below 1 µM in less than 4 d in bothnitrogen-depleted treatments. In the phosphorus-depleted treatments, phosphate concentration after 4d was below 1 µM under axenic conditions, butremained at ≥2 µM under xenic conditions. Phyto-plankton growth in the xenic phosphorus-depletedtreatment may have thus not been phosphorus-lim-ited.

The results from the ANOVAs suggested that bac-terial activity significantly impacted cell growth andthe accumulation of different pools of carbon in nutrient-balanced conditions, but had little overalleffect under nutrient-depleted conditions. Specifi-cally, bacteria promoted diatom growth when grownunder nutrient-balanced conditions, but had littleeffect under nutrient-depleted conditions. In thiscontext, it should be noted that bacterial growth onlyoccurred in the presence of diatom growth, and spe -cifically benefited from optimal growth of the diatoms,suggesting a mutualistic relationship between diatomsand bacteria (Bell & Mitchell 1972, Azam & Ammer-man 1984). It is likely that a trophic interplay be tweenthese organisms exists, more likely of a biotrophicthan of a necro trophic nature.

The DOC patterns observed in the xenic treat-ments of our experiment reflect the net result ofactive release or passive leakage of organic matterfrom diatoms as well as bacterial production, utiliza-tion, and modification of exudates. Marine bacteriarelease exoenzymes that hydrolyze complex DOCmolecules, which may than be taken up by bacteria,supporting their growth (Azam & Cho 1987, Martinezet al. 1996, Grossart et al. 2005, 2007). Under nutrient-balanced conditions, the growth of bacteria re sultedin decreased DOC accumulation (200 µM difference)in the xenic compared to the axenic treatment. Inboth nutrient-depleted treatments, the accumulationof DOC concentration was similar in axenic andxenic treatments, suggesting that bacterial DOC uti-lization was small. Presumably, concentrations ofinorganic nutrients, which are required for bacterialDOC utilization (Ducklow et al. 1986, Carlson &Duck low 1996, Stoderegger & Herndl 1998, Thing -stad et al. 2008) limited bacterial growth rate. How-ever, bacterial cell numbers did in crease in bothnutrient-depleted treatments, albeit less than in thenutrient-balanced treatment. Since DOC release andconsumption occur simultaneously, we cannot ruleout that even in nutrient-depleted treatments bacteriastimulated exudation of organic carbon by Thalas -siosira weissflogii while utilizing the additional DOC.

The most labile components of the marine DOCpool are DFAA and DCAA as well as DFCHO and

DCCOH (Pakulski & Benner 1994, Weiss & Simon1999, Rosenstock & Simon 2001). Pronounced differ-ences in DCAA composition between treatmentsgrown under nutrient-balanced or phosphorus-depleted conditions (irrespective of the presence ofbacteria) suggested that the quality of algal-derivedDOC, or at least the amino acid fraction of DOC,depended largely on the availability of phosphorus.These findings are consistent with previous studies(Myklestad & Haug 1972, Staats et al. 2000, Meon &Kirchman 2001, Granum et al. 2002), which demon-strated that quantity and quality of algal exudationdepend on nutrient availability, through its effect onthe physiological state of the algae. Bacterial activityalso shapes amino acid concentration, e.g. via therelease of enzymes such as alkaline phosphates(Hoppe et al. 1988, Ammerman & Azam 1991, Smithet al. 1992). The highest DFAA and DCAA concen-trations were indeed observed in xenic treatmentsunder phosphorus-depleted conditions. High DFAAconcentration may be the result of extensive DCAAhydrolysis, especially under nutrient-limited condi-tions. This notion is supported by the observationthat in the second half of the experiment, DFAA con-centrations increased while DCAA concentrationsdecreased in the presence of bacteria. Thus a tightcoupling between protein hydrolysis and amino aciduptake can be assumed in our experiments, asdescribed earlier (Hoppe et al. 1988).

Differences in carbohydrate (DCCHO) concentra-tions between nitrogen-depleted treatments andphosphorus-depleted or nutrient-balanced treatmentsimply that nitrogen depletion resulted in increasedDCCHO exudation, but that this was not directlyinfluenced by bacteria.

In our experimental model system, bacteria did notappear to drive DOC composition; rather nutrientavailability—which affected exudation by diatoms—appeared to control quantity and composition of thecarbohydrates and amino acids in the water. Carbonis also partitioned between dissolved and particulatepools. Although the formation of TEP from dissolvedprecursors is largely abiotic (Passow 2000, Verdugo& Santschi 2010), bacteria may contribute to TEP pro-duction (Mari & Kiørboe 1996, Pedrotti et al. 2009).

In our experiments, bacterial activity resulted inincreased TEP abundance and a shift towards largerTEP in the nutrient-balanced treatment, but had nosignificant effect in either of the nutrient-depletedtreatments. Increase in larger TEP size and maximalcell-specific TEP concentration (µm2 diatom−1) occur -red in phosphorus-depleted treatments regardless ofthe presence of bacteria and in the axenic nitrogen-

11

Mar Ecol Prog Ser 461: 1–14, 2012

depleted treatment. The measured TEP concentra-tions observed in the axenic nitrogen-depleted treat-ments were high; more than 2 orders of magnitudelower cell-specific TEP concentrations were ob servedin the xenic nitrogen-depleted treatment, suggestingthat under nitrogen limitation, bacteria utilized TEPextensively. This is in contrast with in situ findings ofincreased TEP concentrations under nitrogen-limitingconditions (Mari & Burd 1998) as well as with studiesof TEP production in other algae cultures undernitrogen-limited conditions (Corzo et al. 2000, Staatset al. 2000, Beauvais et al. 2006).

Cell-surface-bound EPS and EPS fibrils releasedby the diatoms may act as TEP precursors and con-tribute to the TEP pool (Passow & Alldredge 1994a,Zhou et al. 1998). One fraction might be lectin-stainable deoxysugar-rich polysaccharides, whichare known to be well correlated to cell stickiness(Waite et al. 1995) and TEP concentrations, and there-fore important for aggregate formation (Mopper et al.1995). Lectin staining revealed that exopolymers,both those associated with diatom surfaces and thosethat were free, were well developed in the xenic,nutrient-balanced treatment. In this treatment, bac-terial activities may have led to the modification or denovo synthesis of additional polysaccharides whichwere not detected in nitrogen- or phosphorus-de -pleted treatments. We speculate that bacterial colo-nization of algal surfaces and the mutualistic interac-tion under nutrient-balanced conditions may lead tostimulation of excretion of different types of poly -saccharides, which in turn may be important for TEPformation and stickiness. The lack of free stainableEPS material in phosphorus-depleted treatments isnot in line with the observed high cell-specific TEPconcentrations, which suggested that different poly-saccharides were formed with different N:P stoichio -metry non-detectable with the 2 lectins used in thisexperiment.

The results of this study demonstrated distincteffects of the diatom-associated bacterium Marino -bacter ad haerens on (1) diatom growth and (2) algal exudation during nutrient-balanced growth condi-tions (but not during nutrient-depleted conditions).Ni trogen or phosphorus depletion had variable ef -fects on algal exudation. While phosphorus-limitedtreatments differed in the quality and composition ofDCAA, TEP size frequency and re leased polysaccha-rides, nitrogen-limited treatments showed quantita-tive differences in carbohydrates and cell-specificTEP concentrations. It is important to understand thefactors governing the production of exudation prod-ucts, because the quantity and quality of released

DOC and the formation of TEP have the potential tochange the aggregation dynamics of diatoms (All-dredge et al. 1993, Passow 2002b, Bhaskar et al.2005). The aggregation, and subsequent sedimenta-tion, of particulate organic matter are key mechanismsfor oceanic carbon sequestration, and also controlmarine biogeochemical cycles, as well as affectingthe efficiency of the biological carbon pump (Fowler& Knauer 1986, Jahnke 1996).

CONCLUSIONS

In our study, the interplay of the diatom-associatedbacterium Marinobacter adhaerens and the algalmodel species Thalassiosira weissflogii de pended onnutrient stoichiometry and availability. Based on ourdata, we hypothesize that phytoplankton exudationmay have evolved as a mechanism to establish mutu-alism with bacteria, which may shift to commensal-ism when nutrients become limited.

In the field, bacterial and algal populations are farmore complex than in our bilateral model system,and environmental conditions change continuously.Consequently, bacteria−algae interactions in situ arelikely to be very dynamic. However, results obtainedwith our simplistic model system clearly identifynitrogen and phosphorus concentrations and theirstoichiometry as important factors governing bacte-ria−diatom interactions. As the nutrient environmentchanges continuously, for example during the pro-gression of phytoplankton blooms, the type of inter-action between diatoms, their exudates, and bacteriais likely to change continuously. The very dynamicnature of these processes have to be considered andincluded in modeling studies to accurately predictcarbon cycling in the ocean.

Acknowledgements. We thank the members of the AquaticMicrobial Ecology Group at IGB (Leibniz Institute of Fresh-water Ecology and Inland Fisheries) and S. Seebah for technical assistance and fruitful discussions. This work wassupported by the Deutsche Forschungsgemeinschaft (UL169/6-1 and GR 1540-8/1-2) and by the National ScienceFoundation (NSF: OCE-0926711).

LITERATURE CITED

Alldredge AL, Passow U, Logan BE (1993) The abundanceand significance of a class of large, transparent organicparticles in the ocean. Deep-Sea Res I 40:1131−1140

Ammerman JW, Azam F (1991) Bacterial 5 - nucleotidaseactivity in estuarine and coastal marine waters—charac-terization of enzyme activity. Limnol Oceanogr 36:1427−1436

12

Gärdes et al.: Bacterial effects on diatom polymer exudation

Azam F, Ammerman JW (1984) Cycling of organic matter bybacterioplankton in pelagic marine ecosystems: micro -environmental considerations. In: Fasham MJR (ed) Flowsof energy and materials in marine ecosystems. Plenum,New York, NY, p 345−360

Azam F, Cho BC (1987) Bacterial utilization of organic matter in the sea. Symposium of the Society for Gen-eral Microbiology. 4.1. Ecology of microbial communities.Cambridge University Press, Cambridge, p 261−281

Azam F, Malfatti F (2007) Microbial structuring of marineecosystems. Nat Rev Microbiol 5:782−791

Beauvais S, Pedrotti ML, Egge J, Iversen K, Marrase C(2006) Effects of turbulence on TEP dynamics under con-trasting nutrient conditions: implications for aggregationand sedimentation processes. Mar Ecol Prog Ser 323:47−57

Bell W, Mitchell R (1972) Chemotactic and growth responsesof marine bacteria to algal extracellular products. BiolBull (Woods Hole) 143:265−277

Bhaskar PV, Grossart HP, Bhosle NB, Simon M (2005) Pro-duction of macroaggregates from dissolved exopoly-meric substances (EPS) of bacterial and diatom origin.FEMS Microbiol Ecol 53:255−264

Bidle KD, Azam F (1999) Accelerated dissolution of diatomsilica by marine bacterial assemblages. Nature 397:508−512

Bruckner CG, Rehm C, Grossart HP, Kroth PG (2011)Growth and release of extracellular organic compoundsby benthic diatoms depend on interactions with bacteria.Environ Microbiol 13:1052−1063

Carlson CA, Ducklow HW (1996) Growth of bacterioplank-ton and consumption of dissolved organic carbon in theSargasso Sea. Aquat Microb Ecol 10:69−85

Chin WC, Orellana MV, Verdugo P (1998) Spontaneousassembly of marine dissolved organic matter into poly-mer gels. Nature 391:568−572

Corzo A, Morillo JA, Rodriguez S (2000) Production of trans-parent exopolymer particles (TEP) in cultures of Chaeto-ceros calcitrans under nitrogen limitation. Aquat MicrobEcol 23:63−72

Ducklow HW, Purdie DA, Williams PJL, Davies JM (1986)Bacterioplankton: a sink for carbon in a coastal marineplankton community. Science 232:865−867

Fowler SW, Knauer GA (1986) Role of large particles in thetransport of elements and organic compounds throughthe oceanic water column. Prog Oceanogr 16:147−194

Gärdes A, Kaeppel E, Shehzad A, Seebah S and others(2010) Complete genome sequence of Marinobacter ad -haerens type strain (HP15), a diatom-interacting marinemicroorganism. Stand Genomic Sci 3:97−107

Gärdes A, Iversen MH, Grossart HP, Passow U, Ullrich MS(2011) Diatom-associated bacteria are required for ag -gregation of Thalassiosira weissflogii. ISME J 5:436−445

Granum E, Kirkvold S, Myklestad SM (2002) Cellular andextracellular production of carbohydrates and aminoacids by the marine diatom Skeletonema costatum: dielvariations and effects of N depletion. Mar Ecol Prog Ser242:83−94

Grossart HP (1999) Interactions between marine bacteriaand axenic diatoms (Cylindrotheca fusiformis, Nitzschialaevis, and Thalassiosira weissflogii) incubated undervarious conditions in the lab. Aquat Microb Ecol 19:1−11

Grossart HP, Schlingloff A, Bernhard M, Simon M, BrinkhoffT (2004) Antagonistic activity of bacteria isolated fromorganic aggregates of the German Wadden Sea. FEMS

Microbiol Ecol 47:387−396Grossart HP, Levold F, Allgaier M, Simon M, Brinkhoff T

(2005) Marine diatom species harbour distinct bacterialcommunities. Environ Microbiol 7:860−873

Grossart HP, Czub G, Simon M (2006) Algae-bacteria inter-actions and their effects on aggregation and organic matter flux in the sea. Environ Microbiol 8:1074−1084

Grossart HP, Tang KW, Kiørboe T, Ploug H (2007) Compari-son of cell-specific activity between free-living andattached bacteria using isolates and natural assem-blages. FEMS Microbiol Lett 266:194−200

Guillard RRL (1975) Culture of phytoplankton for feedingmarine invertebrates. In: Smith WL, Chanley MH (eds)Culture of marine invertebrate animals. Plenum Press,New York, NY, p 108−132

Hoppe HG, Kim SJ, Gocke K (1988) Microbial decomposi-tion in aquatic environments—combined process ofextracellular enzyme-activity and substrate uptake. ApplEnviron Microb 54:784−790

Jahnke RA (1996) The global ocean flux of particulate or -ganic carbon: aereal distribution and magnitude. GlobalBiogeochem Cycles 10:71−88

Kaeppel EC, Gärdes A, Seebah S, Grossart HP, Ullrich MS(2012) Marinobacter adhaerens sp. nov., a particle asso-ciated bacteria isolated from the German Bight. Int J SystEvol Microbiol 62:124−128

Laspidou CS, Rittmann BE (2002) A unified theory for extracellular polymeric substances, soluble microbialproducts, and active and inert biomass. Water Res 36:2711−2720

Liener I, Sharon N, Goldstein I (1986) The lectins: proper-ties, functions and applications in biology and medicine.Academic Press, London

Logan BE, Passow U, Alldredge AL, Grossart HP, Simon M(1995) Rapid formation and sedimentation of large ag -gregates is predictable from coagulation rates (half-lives)of transparent exopolymer particles (TEP). Deep-Sea ResII 42:203−214

Lovdal T, Eichner C, Grossart HP, Carbonnel V, Chou L,Martin-Jezequel V, Thingstad TF (2008) Competition forinorganic and organic forms of nitrogen and phosphorusbetween phytoplankton and bacteria during an Emilia-nia huxleyi spring bloom. Biogeosciences 5:371−383

Magaletti E, Urbani R, Sist P, Ferrari CR, Cicero AM (2004)Abundance and chemical characterization of extracellu-lar carbohydrates released by the marine diatom Cy -lindrotheca fusiformis under N- and P-limitation. Eur JPhycol 39:133−142

Mari X, Burd A (1998) Seasonal size spectra of transparentexopolymeric particles (TEP) in a coastal sea and com-parison with those predicted using coagulation theory.Mar Ecol Prog Ser 163:63–67

Mari X, Kiørboe T (1996) Abundance, size distribution andbacterial colonization of transparent exopolymeric parti-cles (TEP) during spring in the Kattegat. J Plankton Res18:969−986

Martinez J, Smith DC, Steward GF, Azam F (1996) Variabil-ity in ectohydrolytic enzyme activities of pelagic marinebacteria and its significance for substrate processing inthe sea. Aquat Microb Ecol 10:223−230

Meon B, Kirchman DL (2001) Dynamics and molecular com-position of dissolved organic material during experimen-tal phytoplankton blooms. Mar Chem 75:185−199

Mopper K, Schultz CA, Chevolot L, Germain C, Revuelta R,Dawson R (1992) Determination of sugars in uncon -

13

Mar Ecol Prog Ser 461: 1–14, 2012

centrated seawater and other natural waters by liquid-chromatography and pulsed amperometric detection.Environ Sci Technol 26:133−138

Mopper K, Zhou JA, Ramana KS, Passow U, Dam HG, Dra-peau DT (1995) The role of surface-active carbohydratesin the flocculation of a diatom bloom in a mesocosm.Deep-Sea Res II 42:47−73

Myklestad (1974) Production of carbohydrates by marineplanktonic diatoms. 1. Comparison of 9 different speciesin culture. J Exp Mar Biol Ecol 15:261−274

Myklestad S (1977) Production of carbohydrates by marineplanktonic diatoms. 2. Influence of N-P ratio in growthmedium on assimilation ratio, growth rate, and pro -duction of cellular and extracellular carbohydrates byChaetoceros affinis var. willei (Gran) Hustedt and Skele-tonema costatum (Grev.) Cleve. J Exp Mar Biol Ecol 29:161−179

Myklestad S, Haug A (1972) Production of carbohydrates bythe marine diatom Chaetoceros affinis var. willei (Gran)Hustedt. I. Effect of the concentration of nutrients in theculture medium. J Exp Mar Biol Ecol 9:125−136

Obernosterer I, Herndl GJ (1995) Phytoplankton extra -cellular release and bacterial growth: dependence on theinorganic N:P ratio. Mar Ecol Prog Ser 116:247−257

Pakulski JD, Benner R (1994) Abundance and distribution ofcarbohydrates in the ocean. Limnol Oceanogr 39:930−940

Passow U (2000) Formation of transparent exopolymer parti-cles, TEP, from dissolved precursor material. Mar EcolProg Ser 192:1−11

Passow U (2002a) Production of transparent exopolymerparticles (TEP) by phyto- and bacterioplankton. Mar EcolProg Ser 236:1−12

Passow U (2002b) Transparent exopolymer particles (TEP) inaquatic environments. Prog Oceanogr 55:287−333

Passow U, Alldredge AL (1994a) Abiotic formation of trans-parent exopolymer particles (TEP) from polysaccharidesexcreted by phytoplankton. Abstr Pap Am Chem Soc207:178

Passow U, Alldredge AL (1994b) Distribution, size and bac-terial colonization of transparent exopolymer particles(TEP) in the ocean. Mar Ecol Prog Ser 113:185−198

Pedrotti ML, Beauvais S, Kerros ME, Iversen K, Peters F(2009) Bacterial colonization of transparent exopolymericparticles in mesocosms under different turbulence inten-sities and nutrient conditions. Aquat Microb Ecol 55:301−312

Pierre G, Graber M, Rafiliposon B, Dupuy C, Orvain F, DeCrignis M, Maugard T (2012) Biochemical compositionand changes of extracellular polysaccharides (ECPS)produced during microphytobenthic biofilm development(Marennes-Oléron, France). Microb Ecol 63:157−169

Porter KG, Feig YS (1980) The use of DAPI for identifyingand counting aquatic microflora. Limnol Oceanogr 25:943−948

Rosenstock B, Simon M (1993) Use of dissolved combinedand free amino acids by planktonic bacteria in Lake Constance. Limnol Oceanogr 38:1521−1531

Rosenstock B, Simon M (2001) Sources and sinks of dis-

solved free amino acids and protein in a large and deepmesotrophic lake. Limnol Oceanogr 46:644−654

Sapp M, Gerdts G, Wellinger M, Wichels A (2008) Consum-ing algal products: trophic interactions of bacteria and adiatom species determined by RNA stable isotope prob-ing. Helgol Mar Res 62:283−287

Smith DC, Simon M, Alldredge AL, Azam F (1992) Intensehydrolytic enzyme activity on marine aggregates andimplications for rapid particle dissolution. Nature 359:139−142

Smith DC, Steward GF, Long RA, Azam F (1995) Bacterialmediation of carbon fluxes during a diatom bloom in amesocosm. Deep-Sea Res II 42:75−97

Sonnenschein EC, Gärdes A, Seebah S, Torres-Monroy I,Grossart HP, Ullrich MS (2011) Development of a geneticsystem for Marinobacter adhaerens HP15 involved inmarine aggregate formation by interacting with diatomcells. J Microbiol Methods 87:176−183

Staats N, Stal LJ, Mur LR (2000) Exopolysaccharide produc-tion by the epipelic diatom Cylindrotheca closterium:effects of nutrient conditions. J Exp Mar Biol Ecol 249:13−27

Stal LJ (2003) Microphytobenthos, their extracellular poly-meric substances, and the morphogenesis of intertidalsediments. Geomicrobiol J 20:463−478

Stoderegger K, Herndl GJ (1998) Production and release ofbacterial capsular material and its subsequent utilizationby marine bacterioplankton. Limnol Oceanogr 43:877−884

Strickland JDH, Parsons TR (1972) A practical handbook ofseawater analysis. Fish Res Board Can, Ottawa

Thingstad TF, Bellerby RGJ, Bratbak G, Borsheim KY andothers (2008) Counterintuitive carbon-to-nutrient couplingin an Arctic pelagic ecosystem. Nature 455:387−390

Verdugo P, Santschi PH (2010) Polymer dynamics of DOCnetworks and gel formation in seawater. Deep-Sea Res II57:1486−1493

Verdugo P, Alldredge AL, Azam F, Kirchman DL, Passow U,Santschi PH (2004) The oceanic gel phase: a bridge inthe DOM-POM continuum. Mar Chem 92:67−85

Waite AM, Olson RJ, Dan HG, Passow U (1995) Sugar-containing compounds on the cell surfaces of marinediatoms measured using concanavalin A and flow cytom-etry. J Phycol 31:925−933

Weiss M, Simon M (1999) Consumption of labile dissolvedorganic matter by limnetic bacterioplankton: the relativesignificance of amino acids and carbohydrates. AquatMicrob Ecol 17:1−12

Wigglesworth-Cooksey B, Cooksey KE (2005) Use of fluo-rophore-conjugated lectins to study cell-cell interactionsin model marine biofilms. Appl Environ Microbiol 71:428−435

Zhou J, Mopper K, Passow U (1998) The role of surface-active carbohydrates in the formation of transparentexopolymer particles by bubble adsorption of seawater.Limnol Oceanogr 43:1860−1871

ZoBell CE (1941) Studies on marine bacteria. I. The culturalrequirements of heterotrophic aerobes. J Mar Res 4:42−75

14

Editorial responsibility: Matthias Seaman, Oldendorf/Luhe, Germany

Submitted: January 25, 2012; Accepted: June 21, 2012Proofs received from author(s): July 27, 2012

cite1: cite2: cite3: cite4: cite5: cite6: cite7: cite8: cite9: cite10: cite11: cite12: cite13: cite14: cite15: cite16: cite17: cite18: cite19: cite20: cite21: cite22: cite23: cite24: cite25: cite26: cite27: cite28: cite29: cite30: cite31: cite32: cite33: cite34: cite35: cite36: cite37: cite38: cite39: cite40: cite41: cite42: cite43: cite44: cite45: cite46: cite47: cite48: cite49: cite50: cite51: cite52: cite53: cite54: cite55: cite56: cite57: cite58: cite59: cite60: cite61: cite62: cite63: cite64: cite65: cite66: cite67: cite68: cite69: cite70: cite71: