Vol. 61: 279–289, 2010 AQUATIC MICROBIAL ECOLOGY Published online November 11 doi: 10.3354/ame01447 Aquat Microb Ecol
Contribution to AME Special 4 ‘Progress and perspectives in aquatic microbial ecology’ OPENPEN ACCESSCCESS REVIEW Composition and function of mucilage macroaggregates in the northern Adriatic
Valentina Turk1,*, Åke Hagström2, Nives Kova<1, Jadran Faganeli1
1Marine Biological Station, National Institute of Biology, 6330 Piran, Slovenia 2School of Natural Science, Linnaeus University, 39182 Kalmar, Sweden
ABSTRACT: The episodic hyperproduction of mucilage macroaggregates in the northern Adriatic Sea creates an important site for the accumulation, transformation, and degradation of organic mat- ter. In this review, the structure and function of macroaggregate components in relation to their macrogel and colloidal fractions are discussed. High resolution electron microscopy showed a very complex structure, a honeycomb-like structure of the mucus macroagregates that might grow to macroscopic sizes. The process of the formation and microbial interaction with the physicochemical diversity of the organic matter pool is poorly understood. Whether the in situ bacteria react to the car- bohydrate-rich mucus as an imbalance in its C:N:P ratio or whether the mucus is in fact largely a bac- terial construct in relation to high dissolved organic carbon levels is unknown. The majority of carbo- hydrate and protein macroaggregate pools are potentially degradable, while the great majority of lipids can be preserved in the water column and exported away or finally deposited on the seabed. Our present knowledge indicates that different macroaggregate fractions and components are sub- jected to compositional selective reactivity, with important implications for macroaggregate persis- tence. Future work should reconcile the discrepancies between bacterial ectoenzyme potential activ- ities and biogeochemical degradation sequences based on actual measurements. The determination of biofilm architecture, particularly the spatial arrangement of microcolonies, has profound implica- tions for the function of these complex communities. We need to improve our understanding of the dynamic relationship among bacteria, other microorganisms, and a variety of organic matter forms.
KEY WORDS: Macroaggregates · Microorganisms · Composition · Degradation · Northern Adriatic
Resale or republication not permitted without written consent of the publisher
OCCURRENCE AND ORIGIN (Kaltenböck & Herndl 1992, Malej & Harris 1993, OF MACROAGGREGATES Peduzzi & Weinbauer 1993, Baldi et al. 1997, Azam et al. 1999, Granéli et al. 1999, Herndl et al. 1999, Myk- Macroaggregates of various sizes, colors, and lestad 1999, Najdek et al. 2002, Manganelli & Funari shapes — classified as small flocs, macroflocs, stringers 2003, Degobbis et al. 2005, Kova< et al. 2005). In partic- (1 to 25 cm), clouds (5–10 cm to 3 m), creamy, and ular, diatoms are thought to play a key role in gelatinous surface layers (Stachowitsch et al. 1990) — macroaggregate formation through the release, i.e. occur periodically in the northern Adriatic. Different exudation, of polymeric substances mostly composed ideas have been presented to explain the development of heteropolysaccharides and to a lesser extent of lipids of the mucilage aggregates in the northern Adriatic, and proteins (Pajdak-Stós et al. 2001, Kova< et al. 2002, and a recent overview of these can be found in Giani et 2004, 2008). al. (2005a). In most studies, the importance of phyto- The first report of mucilage appearance in the Adri- plankton for the formation of mucilage has been tested atic dates back to 1729 (Fonda-Umani et al. 1989). In in combination with specific environmental factors, recent decades, the extent and frequency of occur- changes in community structure, nutrient limitation, rence of nuisance mucilage has increased, as has pub- cell lyses or viral attack, and reduced grazing pressure lic discontent (Giani et al. 2005a). The formation of
*Email: [email protected] © Inter-Research 2010 · www.int-res.com 280 Aquat Microb Ecol 61: 279–289, 2010
macroaggregates usually starts in late spring and is high microbial activity in the seawater, the process of thought to be linked to previous changes in the seawa- mucus formation is a result of the combination of ter inorganic N:P ratio rather than to the concentration microbial activity and environmental conditions. Here, of individual nutrients (Cozzi et al. 2004, Penna et al. we intend to present and discuss the role of microor- 2009). The marked retention of fresh water and water ganisms in the formation of mucus aggregates. In order column stratification in the northern Adriatic during to further this understanding, mucilage samples were this period also seems to contribute to the development analyzed using different microscopic techniques. In (Degobbis et al. 2005). In addition, accumulation of addition, the chemical composition of young and age- mucilage organic matter (OM) is enhanced by the ing aggregates is presented in an attempt to describe peculiar hydrological conditions of the northern Adri- the function (activity, role) of microorganisms in mucus atic during summer when this sea region becomes iso- aggregates. lated from the rest of the Adriatic with the formation of a gyre. This change is connected to an increased resi- dence time of the water, development of a pycnocline, STRUCTURE AND CHEMICAL COMPOSITION OF and low turbulent shear (Supic´ & Orlic´ 1999). Estimates MACROAGGREGATES of the volume-specific carbon mass of mucilage-associ- ated suspended matter that accumulated in the water The mucilage appears to be a highly structured column during mucilage events in 1997 and 2000 matrix containing diverse microbial flora, colloidal yielded values of about 50 mg C l–1 for dense water col- organic debris, and inorganic particles. Light and epi- umn mucilage clouds (Malej et al. 2001). According to fluorescent microscopy showed a diverse microbial
Fig. 1. Epifluorescent images of mucus samples in the Gulf of Trieste (northern Adriatic) (30 June 2004). Mucus samples were collected by divers, fixed with formaldehyde, and photographed by Olympus BH2 (DP70 soft). (a,b) DAPI (4’, 6-diamino-2- phenylindole 1 µg ml–1, final Sigma)-stained samples where heterotrophic bacteria of different size and forms are presented (UV filter set). (c) Autofluorescence of numerous cyanobacteria and unicellular Chlorococcus cells within a micro-millimeter mucus aggregate (magnification 200×). (d) Typical nanoflagellate cells with body scales visible after sample staining with DAPI (UV filter set) Turk et al.: Northern Adriatic macroaggregates 281
community attached to/in the mucus (Fig. 1). The most abundant organisms within the macroaggregates were bacteria, cyanobacteria, flagellates, and different diatoms. Different studies have demonstrated a highly variable composition of bacteria and phytoplankton in the mucilage during periods of macroaggregate forma- tion in the northern Adriatic (Rath et al. 1998, Najdek et al. 2002, Kova< et al. 2005, Flander-Putrle & Malej 2008). A large number of samples were examined using transmission electron microscopy (TEM) or scanning electron microscopy (SEM) and showed the complex ul- trastructure of the mucus macroaggregates. SEM analysis, using the high-pressure freezing method (Studer et al. 2008), showed a very complex network at low magnification (Fig. 2a). The inner core of the mucus consisted of fibers forming net-like structures. At higher magnification, it was possible to see fibers of dif- ferent shapes and thickness, organized into complex networks (Fig. 2b,c). In the TEM images of thin mucus layers, the honeycomb-like structure is clearly visible (Fig. 3). A fibrilar, honeycomb structure can grow to macroscopic size in cultures and biofilms, but, as seen in Fig. 3, it also occurs in the free water mass ecosystem in the northern Adriatic. However, the process of macroaggregate formation and how heterotrophic bac- teria interact in the seawater with the great physico- chemical diversity of the OM pool is poorly understood. Physicochemically, macroaggregates represent the transition between colloidal OM (macromolecules), into macrogels (matrix) and particulate OM (POM; Chin et al. 1998, Verdugo et al. 2004, Svetli Fig. 3. Transmission electron microscopy images of the mucus aggregates from the northern Adriatic (30 June 2004). The images of ultra thin layers of mucus show a regular honeycomb structure (Jeol 100 CX electron microscope) The percentages are normally lower in deeper, par- ROLE OF MICROBES IN MACROAGGREGATES tially degraded mucus aggregates (average of total carbohydrate, 7.6%; and total protein, 0.3%). In the pelagic photic zone, nutrient limitation is The macroaggregate interstitial water colloids also believed to be a fundamental controlling factor for the show the presence of both basic constituents, i.e. car- community composition of microorganisms. Malfunc- bohydrates and proteins (Fig. 4), while the bands assigned to lipids, i.e. aliphatic components, are less evident. In all sample analyses, cyanogenic glycosides were detected, which might represent the important part of organic nitrilated compounds, whose origin can be related to amino acids (Legras et al. 1990). The presence of carbohydrates in all colloid fractions was also confirmed by higher UV absorption spectral analyses (at λ = 250 nm; Binkley & Binkley 1999, Giani et al. 2005b). The results of the carbohydrate content and the C:N ratio within the higher molecular weight (MW) colloidal fraction (Fig. 5) suggest that N- containing carbohydrates can be important con- stituents of this fraction. The presence of the non-pro- tein N compounds was also supported by δ15N data exhibiting low values, especially in the >30 kDa frac- tion (Faganeli et al. 2009) in accordance with pub- lished isotopic data of non-protein nitrogen com- Fig. 4. Fourier transform infrared (FTIR) spectra of macro- pounds, for example N-acetylglucosamine (Smucker aggregate colloidal fractions (retentates) with nominal mole- & Dawson 1986), showing that a significant isotope cular weights of 30 to 10 kDa (UF1), 10 to 5 kDa (UF2), depletion of these compounds, namely amino sugars and <5 kDa (UF3); carbohydrate bands (region ~1150 to –1 –1 as well as chlorophyll and nucleic acids, occurs due 900 cm ), protein bands (region 1654 to 1635 cm ), lipid bands (region 2950 to 2850 cm–1), and cyanogenic glycosides to enzymatic transamination (Macko et al. 1986, Sachs (nitrile) band (at 2240 cm–1, assignment after Nishikida & et al. 1999). Hannah 1996) Turk et al.: Northern Adriatic macroaggregates 283 1.2 0.016 ) whether the mucus is in fact largely a –1 Corg content l bacterial construct in relation to high 0.014 g 1 Carbohydrates DOC levels, is unknown. On the other N content tot 0.012 hand, mucus formation in relation to bac- (%) 0.8 0.01 terial growth is well known as, for exam- tot ple, in the case of plaque formation on 0.6 0.008 teeth, biofilm formation on marine sur- 0.006 0.4 faces, and extensively in the carbon-rich 0.004 environment of the pulp and paper pro- (%) and N org 0.2 duction process. In all cases, mucus for- 0.002 C mation is governed by an adaptive bene- 0 0 fit for the bacteria, which suggests that the formation of mucilage macroaggre- UF3/0 UF1/0 UF2/0 UF3/0 UF1/0 UF1/0 UF2/0 UF2/0 UF3/0 UF1/F UF2/F UF1/F UF2/F UF1/F UF2/F Carbohydrate concentration (m gates in the Adriatic should provide a Start of incubation 1st week 4th week benefit for the microorganism that pro- duces the mucus. Fig. 5. Organic carbon (Corg), total nitrogen (Ntot) (Hedges & Stern 1984), and Bacterial cells comprise the largest liv- total carbohydrate (Dubois et al. 1956) contents in macroaggregate colloidal ing surface in marine environments, and fractions (ultrafiltration permeates UF/F and retentates UF/0) with nominal molecular weights of 30 to 10 kDa (UF1), 10 to 5 kDa (UF2) and <5 kDa (UF3) bacterioplankton release DOM to form during 4 wk of incubation at 26°C in the dark gel (Heissenberger et al. 1996, Stodereg- ger & Herndl 1999). The importance of microalgal–bacterial interactions within tioning of the microbial loop has been proposed as a marine snow and the release of extracellular material common feature for the accumulation of dissolved OM have been emphasized previously (Heissenberger et (DOM) in marine systems, wherein P-limited bacteria al. 1996, Leppard et al. 1996). Heissenberger et al. are unable to consume dissolved organic carbon (1996) observed that particle- associated bacteria (DOC) as fast as it is produced (Thingstad et al. 1997). exhibit larger envelopes than free-living bacteria, but Multi-year studies of microbial processes in the coastal the fate of these structures was not studied further. The northern Adriatic Sea showed elevated concentrations microbial species that may form regular structures, of DOC during high-primary-production seasons including ‘honeycombs’ or ‘veils,’ have been studied (Fonda Umani et al. 2007) and the accumulation of mainly in biofilms from cultures or aquatic environ- dissolved and colloidal polysaccharides resistant to ments (Dalton et al. 1996, Trichet et al. 2001, degradation (Azam et al. 1999) as a source material for Schaudinn et al. 2007 and references therein). Mucus mucilage. The release and accumulation of polysac- aggregates, comparable to biofilms, could be defined charides were studied in different controlled enrich- as assemblages of microorganisms and their associated ment experiments, but mainly to understand the re- extracellular products attached to an abiotic or biotic lationship between nutrient limitation, DOM accu- surface (Davey & O’Toole 2000). Biofilm microorgan- mulation, and microbial response (Fajon et al. 1999, isms demonstrate physiological and morphological Malej et al. 2003). Bacterial abundance and production changes during their transition from planktonic growth are considerably higher in the aggregates (Müller- to surface-attached community formation (O’Toole et Niklas et al. 1994) compared to concentrations in al. 2000, Whiteley et al. 2001, Sauer et al. 2002). A surrounding water. During summer, the Adriatic Sea is range of behavior patterns in different surface-coloniz- P-limited, and the dominance of large rod- and vibrio- ing bacteria has been studied, and it is evident that the shaped bacteria has been recorded in some other parts processes are highly regulated. Bacterial cells sur- of the Mediterranean Sea (La Ferla & Leonardi 2005). round themselves to form microcolonies whose shape Recently it was proposed that the growth of bacteria and structure are determined by cell-to-cell signals under nutrient limitation conditions depends on the and are influenced by environmental conditions. Some ‘surface:cell requirement of limiting element’ ratio Gram-negative marine bacteria associated within (Thingstad et al. 2005). This observation fits very well marine snow produce communication signals involved with the response observed for the single organism in quorum sensing, such as acylated homoserine lac- Vibrio splendidus, which increases in size during nutri- tones that are responsible for phenotypic behavior (for ent starvation as a means of gaining better uptake biofilm formation and exoenzyme production) when properties (Løvdal et al. 2008). Whether the in situ bac- the population reaches high densities (Gram et al. teria react to the carbohydrate-rich mucus as an imbal- 2002). There is also recent literature based on genomic ance in the C:N:P ratio like V. splendidus does, or interferences that pelagic bacteria produce surface 284 Aquat Microb Ecol 61: 279–289, 2010 polysaccharides and proteins for interaction with parti- However, we found only a slight increase of C:N cles and organisms (Moran et al. 2007). Recently, Mal- ratios (from 23 to 25, molar) in the water column fatti & Azam (2009) showed that bacteria themselves macroaggregate matrices, suggesting a parallel degra- produced cell-surface architectures that served as the dation (similar degradation kinetics) of macroaggre- structural basis for the networks at nanometer to gate carbohydrates and proteins in the summer strati- micrometer scales. High-resolution atomic force fied seawater column (Posedel & Faganeli 1991). This microscopy analyses showed the interconnection of finding differs from those reported by others in the cell-surface gel matrices between microbial cells. The northern Adriatic, i.e. significant increasing C:N ratios phylogenic specificities of microbes, their high abun- in older macroaggregates (Del Negro et al. 2005). Our dance, and potential for aggregation to form complex observation was further supported by Fourier trans- networks are confirmed as quantitatively significant form infrared (FTIR) spectroscopy analyses showing processes and may influence the biochemistry of the the presence of carbohydrates and proteins in surface mucus macroaggregates. and water column samples while lipid contents were lower (Fig. 6). The selective degradation of bulk macroaggregate carbohydrates is reflected in the neu- MACROAGGREGATE DEGRADATION tral monosaccharide composition showing decreasing glucose content with macroaggregate age (Faganeli et In this review, we assume that the mucilage forms as al. 1995, Giani et al. 2005b). the result of an unbalanced C:N:P ratio in the water col- According to Corg, Ntot (Fig. 5), and high perfor- umn, which can favor extensive mucus structures that mance size-exclusion chromatography (N. Koron et al. would benefit the organism in some specific way. This unpubl.) analyses, the degradation of macroaggregate also means that the mucus must have a lifecycle initiated colloidal fractions proceeds faster in the >30 kDa and and growing as long as the microbes can take advantage 30 to 10 kDa fractions compared to the 10 to 5 kDa of the mucus environment. With changing environmen- fraction. The differences in reactivity can again be ex- tal conditions, the mucus environment can no longer of- plained by higher levels of N-containing polysaccha- fer a selective advantage, and thus the mucus starts to rides in the higher MW fractions and rapid degradation become old and a degradation process is initiated. of carbohydrates in all studied fractions leading to the Bacteria are the principal decomposers of biopoly- preservation of organic nitrogen in aged degraded mers in macroaggregates (Müller-Niklas et al. 1994, (‘mature’) macroaggregates. Recent reports under- Mingazzini & Take 1995). Macroaggregate degrada- score up to a 3-fold faster microbial degradation of the tion is supported by high bacterial abundance and carbohydrate component of glycoproteins compared to activity, and by an efficient recycling of nutrients, the proteinaceous component (Ogawa et al. 2001, which seem dependent on aggregate type and age Nagata et al. 2003). (Del Negro et al. 2005). Bacterial extracellular enzymes are important catalysts in the degradation of macro- aggregate OM to DOM, significantly influencing the marine biogeochemical cycling of organic carbon and other elements that limit microorganism growth. The formed low MW products (<600 Da) can be taken up by heterotrophic prokaryotes (Nagata 2000, Conan et al. 2007) and are substrates for subsequent transforma- tions including microbial and photochemical degrada- tion, and polymerization (Kova< et al. 1998). The qual- ity of OM, including its C, N, P stoichiometric balance, determines the proportion between biomass produc- tion and remineralization (Obernosterer & Herndl 1995, del Giorgio & Cole 2000). Moreover, the transfor- mation and degradation of macroaggregate OM encompass size- and compositional-dependent reac- tivity changes (Benner 2002). The degradation of sink- ing macroaggregates, enriched in the interstitial water oxygen concentration, is dependent on their residence Fig. 6. Fourier transform infrared (FTIR) spectra of surface and water column macroaggregate matrix; carbohydrate time (Ploug et al. 1999), which is connected to the strat- bands (region ~1150–900 cm–1), protein bands (region ification of the northern Adriatic seawater column and 1654–1635 cm–1), lipid bands (region 2950–2850 cm–1), and the formation of the summer gyre (Supi´c & Orli´c 1999). inorganic (mineral) components (region <1000 cm–1) Turk et al.: Northern Adriatic macroaggregates 285 ENZYMATIC HYDROLYSIS OF 15 a C M MA MACROAGGREGATES ) –1 Bacteria associated with macroaggregates exhibit a 10 very high potential enzymatic hydrolysis compared to free-living cells in the surrounding water (Del Negro 5 et al. 2005, Zoppini et al. 2005), and high potential aminopeptidase activity suggests that proteins are de- 0 graded more rapidly than other constituents (Simon et 01234567 al. 2002), which is in accordance with the reported pref- )b TCHO (µg ml erential degradation of organic N compared to C in ag- –1 1.5 gregates (Grossart & Ploug 2001) and marine snow h –1 (Müller-Niklas et al. 1994). This can be reflected in a 1 progressive increase of carbon content in the macroag- gregates. The potential activity of lipase, one of the most active ectoenzymes of aquatic bacteria (Zoppini et 0.5 BCP (µgC l al. 2005), was also found to be very effective in macroaggregates. Glucosidase activity was reported to 0 be lower compared to aminopeptidase activity. It was 01234567 also proposed that the rapid hydrolysis of proteina- ) 35 c –1 l ceous material implies its large availability for bacteria 9 30 (Zoppini et al. 2005). However, Danovaro et al. (2005) 10 25 found high β-glucosidase activity in the northern Adri- x 20 atic waters in summer during a mucilage event. The lower contribution of potential polysaccharide hy- 15 drolytic activity can lead to enrichment in polysaccha- 10 ride content presumably rich in refractory compounds. 5 β The most detectable -glucosidase activity can indicate Bacteria (cells 0 the shift towards a more refractory β-glucosidic bond in 01234567 aged macroaggregates (Müller-Niklas et al. 1994). The Time (d) potential hydrolysis of organophosphorus compounds Fig. 7. (a) Concentrations of total dissolved carbohydrates by alkaline phosphatase indicated slower P cycling (TCHO), (b) bacterial carbon production (BCP), and (c) bac- compared to N (peptidase activity) and, thus, an in- terial abundance during the macroaggregate degradation creasing bacterial P limitation in aged macroaggre- experiment (C: control, M: addition of bacterial isolates, MA: gates (Zoppini et al. 2005). addition of bacterial isolates and inorganic N, P nutrients). TCHO were measured according to the protocol of Dubois This ‘spectrum’ of potential extracellular enzymatic et al. (1956), BCP by the 3H-leucine incorporation and centri- activities was not directly reflected in the chemical fugation method (Smith & Azam 1992), and bacterial abun- composition of the progressive aged macroaggregates dance with DAPI staining (Porter & Feig 1980). Error bars or in laboratory degradation experiments. During the are ±SD incubation of macroaggregates, carbohydrates and proteins decreased rapidly, but a slower decrease of higher bacterial uptake evidenced by increasing bac- lipids, only at the end of the experiment, was presum- terial biomass and bacterial carbon production. The ably due to lower lipid degradability (Fig. 6). This simultaneous addition of inorganic N and P nutrients observation was in agreement with previously re- had a negligible effect on bacterial density and pro- ported studies using 1H-nuclear magnetic resonance duction, suggesting that bacterial C-uptake should not (NMR) spectroscopy performed on bulk macroaggre- be severely limited by inorganic nutrients, and exhib- gates collected in situ, showing that carbohydrates ited a rather limited effect on carbohydrate release. degrade much faster than lipids (Kova< et al. 2002). The study of laboratory-based degradation of macroaggregate carbohydrates by the addition of cul- ‘BIOAVAILABLE’ MACROAGGREGATE POOL tured bacterial strains, isolated from the Gulf of Tri- este, in parallel with bacterial cell counting and bacte- The bulk macroaggregates from the Gulf of Trieste rial production revealed a rapid increase of dissolved (2004) contained about 14% carbohydrates and 5% pro- carbohydrate concentrations followed by a slow teins (Penna et al. 2009). The enzymatically hydrolyzed decrease (Fig. 7). The latter was probably due to carbohydrates and proteins comprised about 70 and 286 Aquat Microb Ecol 61: 279–289, 2010 >50%, respectively, of carbohydrate and protein pools in a phytoplankton bloom are of enormous interest to surface and water column macroaggregates. Both con- understand the role of dissolved and colloidal OM and stituents, probably bonded into glycoproteins, can be de- its aggregation to form mucilage under relevant eco- graded in parallel. The percentages were similar to those system conditions. The chemical composition of such estimated from chemical analyses of particulate and sed- aggregates is influenced by various factors, including imenting OM containing macroaggregates in the sum- nutrients, temperature, pH, physiology, and age of the mer stratified seawater column in the Gulf of Trieste aggregate. The available information is inadequate to (Posedel & Faganeli 1991). The bacterial degradation explain the origin/mechanisms of mucus aggregate of macroaggregate polysaccharides studied in vitro formation in both laboratory and field studies. Micro- confirmed the rapid release of dissolved carbohydrates bial extracellular polymeric substances as such are and later a slowed decrease, probably due to higher widely distributed in marine environments and act as bacterial activity. The remaining less-degradable poly- hotspots for the transformation of OM and modification saccharides, containing more β-glycosidic linkages, of the micro-environment. Laboratory studies of these and low-degradable proteinaceous matter significantly microenvironments are few, and lack of appropriate contribute to macroaggregate persistence. This indicates tools restricts in situ studies of various processes occur- that the degradation, persistence, and accumulation of ring in these micro-niches. According to the complex- macroaggregates are biochemically fractionated and ity of the mucus material, a combination of techniques could not simply be the result of rather low bacterial glu- is required to access the different parameters which cosidase activity observed in the northern Adriatic (Za- contribute to mucilage formation: organic and inor- ccone et al. 2002). Considering the macroaggregate lipid ganic compounds (molecules) and microorganisms. content of about 2% (Penna et al. 2009), the hydrolyzed Recent and continuous methodological advances in lipids should comprise only about 30% of the total lipid marine microbiology will allow better imaging and pool. Hence, most of the macroaggregate lipids includ- analytical techniques to determine structure on a ing glycolipids, contrary to more degradable carbohy- (sub)micrometer scale, with non-destructive methods. drates and proteins, would persist in the summer strati- The determination of biofilm architecture, particularly fied water column (Kova< et al. 2002) and would finally the spatial arrangement of microcolonies, might have be deposited on the seabed and transported southward important implications for the function of these com- into the deeper basin (Fonda Umani et al. 2007). The la- plex biochemical structures. bility of the macroaggregate colloidal fraction is also de- Another area that has not been explored much is the pendent on its composition. Since the polysaccharide change in bacterial diversity during aggregation and component in the lower MW colloidal fraction seems degradation of macroaggregates. Intensive studies more degradable compared to N-containing polysaccha- using new molecular techniques need to be carried ridic material, the higher MW fraction represents a pos- out to assess the microbial diversity and reveal the sible path of organic nitrogen preservation in marine changes in microbial community during the production organic colloids. As a concurrent process, the photo- and degradation of mucus macroaggregates. Future chemical degradation of macroaggregate colloids work should also be focused on bacterial ectoenzyme cleaves glycosidic bonds, producing oligomers and activities and biogeochemical degradation processes monomers (Kova< et al. 1998). All of these aspects can within the water column as well as above the bottom, also have an important impact on metal, for example Hg, since the mucus macroaggregates represent a danger and OM interactions leading to metal immobilization as to seafloor communities (Schiaparelli et al. 2007). well as its release into the seawater medium (Guo & Santschi 2007). Future work should reconcile the dis- Acknowledgements. This research was financed by the crepancies between observed macroaggregate ectoen- Slovene Research Agency (P1-0237, J1-2136). We are grateful zyme potential activities and biogeochemical degrada- to F. Tatti (FEI com Italy), L. Bonzi (University of Florence), tion sequences and enzyme hydrolysis. This difference and M. Heldal (Marine Microbiology, Institute for Biology, University of Bergen) for providing assistance with electron underlines the important relationship between substrate microscopy, and anonymous reviewers for their critical and composition and its lability. valuable comments on the manuscript. LITERATURE CITED CONCLUDING REMARKS Azam F, Fonda Umani S, Funari E (1999) Significance of bac- Mucilage is a compact and dense material, with a teria in the mucilage phenomenon in the northern Adriatic Sea. Ann Ist Super Sanit 35:411–419 huge heterogeneity in its network, chemical, and Baldi F, Minacci A, Saliot A, Mejanelle L, Mozetic P, Turk V, organism composition and structures. Conformation Malej A (1997) Cell lysis and release of particulate poly- changes in macromolecules and their complexes after saccharides in extensive marine mucilage assessed by Turk et al.: Northern Adriatic macroaggregates 287 lipid biomarkers and molecular probes. Mar Ecol Prog Ser Fonda Umani S, Del Negro P, Larato C, De Vittor C and oth- 153:45–57 ers (2007) Major inter-annual variations in microbial Benner R (2002) Chemical composition and reactivity. In: dynamics in the Gulf of Trieste (northern Adriatic Sea) and Hansell DA, Carlson CA (eds) Biogeochemistry of marine their ecosystem implications. Aquat Microb Ecol dissolved organic matter. Elsevier, Amsterdam, p 367–404 46:163–175 Binkley ER, Binkley RW (1999) Carbohydrate photochemistry. Giani M, Degobbis D, Rinaldi A (2005a) Mucilages in the American Chemical Society, Washington, DC Adriatic and Tyrrhenian Seas. Sci Total Environ 353:1–2 Chin WC, Orellana MV, Verdugo P (1998) Spontaneous Giani M, Berto D, Zangrando V, Castelli S, Sist P, Urbani R assembly of marine dissolved organic matter in polymer (2005b) Chemical characterization of different typologies gels. Nature 391:568–572 of mucilaginous aggregates in the Northern Adriatic Sea. Clark AH, Farrer DB (1995) Kinetics of biopolymer gelation — Sci Total Environ 353:232–246 implications of a cascade theory description for the con- Gram L, Grossart HP, Schlingloff A, Kiørboe T (2002) Possible centration, molecular weight, and temperature depen- quorum sensing in marine snow bacteria: production of acy- dences of the shear modulus and gel time. J Rheol 39: lated homoserine lactones by Roseobacter strains isolated 1429–1444 from marine snow. Appl Environ Microbiol 68: 4111–4116 Conan P, Sondergaard M, Kragh T, Thingstad F and others Granéli E, Carlsson P, Turner JT, Tester PA, Béchemin C, (2007) Partitioning of organic production in marine plank- Dawson R, Funari E (1999) Effects of N:P:Si ratios and zoo- ton communities: the effects of inorganic nutrient ratios plankton grazing on phytoplankton communities in the and community composition on new dissolved organic northern Adriatic Sea. I. Nutrients, phytoplankton bio- matter. Limnol Oceanogr 52:753–765 mass, and polysaccharide production. Aquat Microb Ecol Cozzi S, Ivan Leppard GG, Heissenberger A, Herndl GJ (1996) Ultrastruc- Penna N, Kova< N, Ricci F, Penna A, Capellacci S, Faganeli J ture of marine snow. I. Transmission electron microscopy (2009) The role of dissolved carbohydrates in the northern methodology. Mar Ecol Prog Ser 135:289–298 Adriatic macroaggregate formation. Acta Chim Slov Løvdal T, Skjoldal EF, Heldal M, Norland S, Thingstad TF 56:305–314 (2008) Changes in morphology and elemental composition Ploug H, Grossart HP, Azam F, Jørgensen BB (1999) Photo- of Vibrio splendidus along a gradient from carbon-limited synthesis, respiration, and carbon turnover in sinking to phosphate-limited growth. Microb Ecol 55:152–161 marine snow from surface waters of Southern California Macko SA, Estep MLF, Engel MH, Hare PE (1986) Kinetic Bight: implications for the carbon cycle in the ocean. Mar fractionation of stable nitrogen isotopes during amino acid Ecol Prog Ser 179:1–11 transamination. Geochim Cosmochim Acta 50:2143–2146 Porter KG, Feig YS (1980) The use of DAPI for identifying and Malej A, Harris RP (1993) Inhibition of copepod grazing by counting aquatic microflora. Limnol Oceanogr 25:943–948 diatom exudates: a factor in the development of mucus Posedel N, Faganeli J (1991) Nature and sedimentation of aggregates? Mar Ecol Prog Ser 96:33–42 suspended particulate matter during density stratification Malej A, Petelin B, Cermelj B (2001) Quantification of in shallow coastal waters (Gulf of Trieste, northern Adri- mucilage-associated suspended matter in the Gulf of atic). Mar Ecol Prog Ser 77:135–145 Trieste (Adriatic Sea). Ann Ser Hist Nat 11:43–52 Raghavan SR, Cipriano BH (2005) Gel formation: phase Malej A, Mozeti< P, Turk V, Terzic ´ S, Ahel M, Cauwet G diagrams using tabletop rheology and calorimetry. In: (2003) Changes in particulate and dissolved organic mat- Weiss RG, Terech P (eds) Molecular gels: materials with ter in nutrient-enriched enclosures from an area influ- self-assembled fibrillar networks. Springer, Amsterdam, enced by mucilage: the northern Adriatic Sea. J Plankton p 233–244 Res 25:949–966 Rath J, Wu KY, Herndl GJ, DeLong EF (1998) High phylo- Malfatti F, Azam F (2009) Atomic force microscopy reveals genetic diversity in a marine-snow-associated bacterial microscale networks and possible symbioses among assemblage. Aquat Microb Ecol 14:261–269 pelagic marine bacteria. Aquat Microb Ecol 58:1–14 Sachs JP, Repeta DJ, Goercke R (1999) Nitrogen and carbon Manganelli M, Funari E (2003) Le mucillagini nel Mar Adri- isotopic ratios of chlorophyll from marine phytoplankton. atico: ruolo dei possibili agenti causali e dei fattori ambi- Geochim Cosmochim Acta 63:1431–1441 entali. Ann Ist Super Sanit 39:77–95 Sauer K, Camper AK, Ehrlich GD, Costerton JW, Davies DG Mingazzini M, Take B (1995) Summary and conclusions of the (2002) Pseudomonas aeroginosa displays multiple pheno- workshop on marine mucilages in the Adriatic Sea and types during development as a biofilm. J Bacteriol 184: elsewhere. Sci Total Environ 165:9–14 1140–1154 Moran MA, Belas R, Schell MA, González JM and others Schaudinn C, Stoodley P, Kainovic ´ A, O’Keeffe T and others (2007) Ecological genomics of marine roseobacters. Appl (2007) Bacterial biofilms, other structures seen as main- Environ Microbiol 73:4559–4569 stream concepts. Microbe 2:231–237 Müller-Niklas S, Schuster S, Kaltenböck E, Herndl GJ (1994) Schiaparelli S, Castellano M, Povero P, Sartoni G, Cattaneo- Organic content and bacterial metabolism in amorphous Vietti R (2007) A benthic mucilage event in North-Western aggregations of the Northern Adriatic Sea. Limnol Mediterranean Sea and its possible relationships with the Oceanogr 39:58–68 summer 2003 European heatwave: short-term effects on Myklestad SM (1999) Phytoplankton extracellular production littoral rocky assemblages. PSZN I: Mar Ecol 28:341–353 and leakage with considerations on the polysaccharide Simon M, Azam F (1989) Protein content and protein synthe- accumulation. Ann Ist Super Sanit 35:401–404 sis rates of planktonic marine bacteria. Mar Ecol Prog Ser Nagata T (2000) Production mechanisms of dissolved organic 51:201–213 matter. In: Kirchman DL (ed) Microbial ecology of the Simon M, Grossart HP, Schweitzer B, Ploug H (2002) Micro- oceans. Wiley-Liss, New York, NY, p 121–152 bial ecology of organic aggregates in aquatic ecosystems. Nagata T, Meon B, Kirchman DL (2003) Microbial degrada- Aquat Microb Ecol 28:175–211 tion of peptidoglycan in seawater. Limnol Oceanogr Smucker RA, Dawson R (1986) Products and photosynthesis 48:745–754 by marine phytoplankton: chitin in TCA ‘protein’ precipi- Najdek M, Degobbis D, Miokovic ´ D, Ivan Thingstad TF, Øvreås L, Egge JK, Løvdal T, Heldal M (2005) Whiteley M, Bangera MG, Bumgarner RE, Parsek MR, Use of non-limiting substrates to increase size: a generic Teitzel GM, Lory S, Greenberg EP (2001) Gene expres- strategy to simultaneously optimize uptake and minimize sion in Pseudomonas aeroginosa biofilms. Nature 413: predation in pelagic osmotrophs? Ecol Lett 8:675–682 860–864 Trichet J, Défarge C, Tribble J, Tribble G, Sansone F (2001) Zaccone R, Caruso G, Cal C (2002) Heterotrophic bacteria in Christmas Island lagoonal lakes, models for the deposition the northern Adriatic Sea: seasonal changes and ecto- of carbonate–evaporite–organic laminated sediments. enzyme profile. Mar Environ Res 54:1–19 Sediment Geol 140:177–189 Zoppini A, Puddu A, Fazi S, Rosati M, Sist P (2005) Extracel- Verdugo P, Alldredge AL, Azam F, Kirchman DL, Passow U, lular enzyme activity and dynamics of bacterial commu- Santschi PH (2004) The oceanic gel phase: a bridge in the nity in mucilaginous aggregates of the northern Adriatic DOM–POM continuum. Mar Chem 92:67–85 Sea. Sci Total Environ 353:270–286 Submitted: April 12, 2010; Accepted: August 12, 2010 Proofs received from author(s): October 20, 2010