Acta Botanica 2-2015.indd ACTA BOT. CROAT. 74 (2), 2015 407 Acta Bot. Croat. 74 (2), 407–422, 2015 CODEN: ABCRA 25 ISSN 0365-0588 eISSN 1847-8476 DOI: 10.1515/botcro-2015-0030 Colonization of diatoms and bacteria on artifi cial substrates in the northeastern coastal Adriatic Sea MAJA MEJDANDŽIĆ1*, TOMISLAV IVANKOVIĆ1, MARTIN PFANNKUCHEN2, JELENA GODRIJAN2, DANIELA MARIĆ PFANNKUCHEN2, JASNA HRENOVIĆ1, ZRINKA LJUBEŠIĆ1 1 University of Zagreb, Faculty of Science, Department of Biology, Roosevetov trg 6, 10000 Zagreb, Croatia 2 Ruđer Bošković Institute, Centre for Marine Research, G. Paliaga 5, 52210 Rovinj, Croatia Abstract – Every surface that is immersed in seawater becomes rapidly covered with an unavoidable biofi lm. Such biofi lm formation, also known as fouling, is a complex multi- stage process and not yet thoroughly investigated. In this study, the succession of diatoms and bacteria was investigated during a one month exposure on an artifi cial substrate of plexiglass (polymer of methyl methacrylate) mounted above the seafl oor at a depth of 5 m. For biofi lm analyses, the fouling was investigated using selective agar plates, epifl uores- cence, light and electronic microscopy, as well as high performance liquid chromatogra- phy (HPLC) pigment analysis. During biofi lm development, the abundance of all biofi lm components increased and reached maximum values after a one month exposure. In the bacterial community, heterotrophic marine bacteria were dominant and reached 1.96 ± 0.79 × 104 colony forming units (CFU) cm–2. Despite the fact that faecal coliforms and intes- tinal enterococci were detected in the water column, faecal coliforms were not detected in the biofi lm and intestinal enterococci appeared after one month of exposure but in the neg- ligible number of 60 ± 10 CFU cm–2. The phototrophic component of the biofi lm was domi- nated by diatoms and reached a concentration of 6.10 × 105 cells cm–2, which was support- ed by pigment analysis with fucoxanthin as dominant pigment in a concentration up to 110 ng cm–2. The diatom community was dominated by Cylindrotheca closterium and other pennate benthic diatoms. A detailed taxonomic analysis by electronic microscopy revealed 30 different taxa of diatoms. The study confi rmed that a plexiglass surface in a marine en- vironment is susceptible to biofouling within 30 days of contact. Furthermore, the co loni- zation process sequence fi rstly involved bacteria and cyanobacteria, and secondly diatoms, which together formed a primary biofi lm in the sea. Keywords: bacteria, biofi lm, biofouling, diatoms, succession * Corresponding author, e-mail: mmejdandzic@gmail.com MEJDANDŽIĆ M., IVANKOVIĆ T., PFANNKUCHEN M., GODRIJAN J., MARIĆ PFANNKUCHEN D., et al. 408 ACTA BOT. CROAT. 74 (2), 2015 Introduction A biofi lm is an assemblage of adhered cells and their products on a surface, a ‘coating’ or ‘covering’ composed of organisms like bacteria, protozoa, algae and invertebrate ani- mals (O’TOOLE et al., 2000, STOODLEY et al., 2002). As a general rule, biofi lm growth can be explained in several phases: (i) adsorption – binding of dissolved chemical compounds, macromolecules such as glycoproteins, polysaccharides and proteoglycans, in the fi rst mo- ments of contact with any surface immersed in the seawater, (ii) immobilization – reversi- ble binding of bacterial cells with weak links and interactions on the surface of the sub- strate, (iii) consolidation – irreversible binding of bacterial cells on the surface of the substrate in which bacterial cells begin to secrete extracellular polymeric substances (EPS), which creates a permanent bond between cells and surfaces, (iv) settling – the fi nal stage in the development of biofi lms, when other micro-organisms such as unicellular algae inhabit bacterial colonies and biofi lm takes a three-dimensional structure. Colonization of microor- ganisms on the solid surfaces is a worldwide problem: from offshore oil platforms and bridges that can collapse due to biofouling, through freighters, cruisers and naval vessels to fi shing facilities and a variety of fi shing gear. The biggest problem of fouling on ships is the possibility of corrosion of the stern, which increases ships fuel consumption up to 30% (DE RINCON et al. 2001). Consequently, there is an increased release of greenhouse gases contri- buting to the major environmental problem of our time. Because of fouling on aquaculture installations for fi sh, shellfi sh and other organisms their cultivation is faced with environ- mental problems such as anoxia, eutrophication, increased turbidity, and all of these can lead to plague organisms and major economic losses (LEWIS et al. 1997). Settling allows for the colonization of multicellular organisms such as larvae, invertebrates, multicellular fi la- mentous algae and other macro-invertebrates (LEHAITRE and COMPÈRE 2007). Periphyton includes all plant and animal organisms attached to various types of substrates that are sub- merged in water, and that do not penetrate the surface and is divided into ‘euperiphyon’ (basic part of the periphyton, formed by attached organisms adapted to a sessile lifestyle) and ‘pseudoperiphyton’ (part that is associated with the periphyton formed by communities of organisms that move freely among attached species, depending on them as a source of food and protection from predators and planktonic organisms caught and retained in a dense network of organic matter). Investigations of biofi lm formation in the marine environment are scarce. It is known that colonization of artifi cial differs from that of natural substrates (HAMILTON and DUTHIE 1984, SABATER et al. 1998). Periphyton communities in the northern Adriatic consist mainly of diatoms with a distinct seasonal variation. The highest abundance and biomass were ob- served in the period between February and October (557.156 ± 82.602 cells cm–2), while in the period between January and February abundance and biomass was much lower (365 ± 407 cells cm–2) (TOTTI et al. 2007). Also, no signifi cant difference in the periphyton com- munity structure and composition were observed in different artifi cial substrates. Investiga- tion of periphyton development on artifi cial substrates in the highly stratifi ed estuary of the karstic Zrmanja River (middle Adriatic) (CAPUT et al. 2008), showed high diatom abun- dance after a two-week exposure (2.3 × 107 cells cm–2) with species richness 41, while after 4 weeks, abundance doubled and richness increased to 50. Periphyton was composed most- ly of Amphora coffeaeformis and Navicula veneta after 2 weeks; while after 4 weeks, Melo- sira moniliformis was co-dominant. On the other hand, ROMAGNOLI et al. (2007) described the diatom communities associated with Eudendrium racemosum as a natural substrate COLONIZATION OF DIATOMS AND BACTERIA IN THE NE ADRIATIC ACTA BOT. CROAT. 74 (2), 2015 409 throughout an annual cycle in the Ligurian Sea. Diatom abundance values ranged from 46.752 ± 24.684 cells mm–2 in February 2003 to 917 ± 331 cells mm–2 in October 2003, while biomass ranged from 1.94 ± 1.94 to 0.013 ± 0.003 μg C mm–2 for the same periods. In contrast, fi lamentous cyanobacteria appeared with high densities between late spring and summer with maximum abundance of 29.872 ± 4.482 cells mm–2 and biomass of 0.32 ± 0.18 μg C mm–2. On average, motile diatoms represented the most abundant fraction of dia- tom communities (73%), followed by adnate (17%), erect (7%) and tube-dwelling growth forms (3%). Considering biomass, motile diatoms represented 48% of the total biomass value, followed by erect (25%), adnate (18%) and tube-dwelling diatoms (9%) (ROMAGNOLI et al. 2007). The aim of this study was to investigate the succession and settling of benthic micro- algae in the process of formation of primary biofi lm in the northeast Adriatic Sea. An addi- tional importance of this research is the interdisciplinary approach in which various meth- ods were combined to obtain a better insight into the biofi lm formation. Thus, the objectives of this study were to (i) determine the quantitative and qualitative composition of diatoms and bacteria on artifi cial plexiglass plates, (ii) to demonstrate their distribution through the investigated time of 30 days and (iii) to provide insight in their probable mutual infl uence on the growth and development of the biofi lm. Materials and methods Sampling Periphyton sampling was carried out in the bay Val de Lesso in the city of Rovinj (45.060N, 13.3748E) in the period from September 9th to October 7th 2013. The bay is un- der moderate anthropogenic infl uence due to the surrounding inhabited area. It is shallow (maximum depth up to 10 m) and the seabed is covered by seagrass Cymodocea nodosa. The late summer/early autumn was chosen for the exposure period according to HILLE- BRAND and SOMMER (1997) due to the appropriate temperature for the endorsement of speed of settlement on artifi cial substrates. Continuous weather measurements were conducted at a meteorological station located at Sv. Ivan na Pučini, Rovinj, western Istrian coast (45.04746N, 13.62167E). The data were supplied by the Croatian Meteorological and Hy- drological Service (Tab. 1). All data regarding abundance of bacteria and diatoms were sta- tistically analyzed by software STATISTICA 7.0 and 2007 Microsoft Offi ce Suite Service Pack 3. Plexiglass was used as a substrate for biofi lm formation following the recommendation of SLÀDEČKOVÀ (1962a) because of its convenience compared to a natural substrate. Dimen- sions of plates were 70 × 20 × 2 mm (l/w/h) for bacteriological sampling and 100 × 70 × 2 mm (l/w/h) for algological sampling. The plates were set in a steel structure with plastic rails (Fig. 1a, 1b). In an effort to refi ne the terminology SLÀDEČKOVÀ (1962a) discovered periphyton exclusive of epiphytic and epizoic forms. Periphyton was subdivided into ‘eu- periphyon’, immobile organisms attached to the substrate by means of rhizoides, gelatinous stalks, or other holdfast mechanisms, and ‘pseudoperiphyton’ forms that are free-living, mobile, or creeping among or within the euperiphyton. The plates were set vertically in or- der to reduce the accumulation of sludge and biomass of ‘pseudoperiphytic’ diatoms (SLÀDEČKOVÀ 1962a). The construction was set in the sea at a depth of 5 m in the meadow of seagrass Cymodocea nodosa and raised 30 cm above the sediment. The depth of fi ve MEJDANDŽIĆ M., IVANKOVIĆ T., PFANNKUCHEN M., GODRIJAN J., MARIĆ PFANNKUCHEN D., et al. 410 ACTA BOT. CROAT. 74 (2), 2015 meters ensures enough light for biofi lm development, while benthic impacts of waves and tides do not interfere with the succession. Two large and two small plexiglass plates were placed into the construction for each sample with a total of 24. Plexiglas plates were previ- ously disinfected with 70% ethanol and stored in sterile containers. The collected material for diatom analyses was preserved in 4% formaldehyde. Bacteriological analysis For bacteriological analysis, plates were removed from the structure after 1 h, 12 h, 24 h, one week and one month of contact. Plates were gently washed with 100 mL of sterile saline (0.8% NaCl) and immersed in Schott bottles containing 150 mL of sterile saline solu- tion so that the entire surface of the plate was immersed in liquid. Biofi lm was dispersed using an ultrasonic probe (40 W, 120 one second cycles). After treatment with ultrasound Tab. 1. Hydrographic parameters of sampling events at Bay Val de Lesso, Rovinj in the period from September 9th to October 7th 2013, Avg. – average value, wind strength is shown in values according to Beaufort scale (0–12). Parameters / Date 9-Sept 10-Sept 11-Sept 13-Sept 16-Sept 7-Oct Avg. daily temperature (°C) 24.2 23.1 18.3 17.3 18.6 14.2 Avg. sea temperature (°C) 22.20 22.00 22.00 22.20 22.00 19.90 Daily pressure (hPa) 1014.4 1013.3 1011.3 1016.9 1003.7 1020.9 Precipitation (mm) 0.3 0.0 0.2 0.5 48.9 20.8 Wind direction SSE SE SE ESE NE ENE Wind strength 4 1 1 1 3 1 Fig. 1. Construction with plexiglass plates placed in meadow of Cymodocea nodosa seagrass (a), plexiglass plates placed in steel structure with plastic rails (b). COLONIZATION OF DIATOMS AND BACTERIA IN THE NE ADRIATIC ACTA BOT. CROAT. 74 (2), 2015 411 supernatant, samples were collected and inoculated on the appropriate culture medium in duplicates. Dispersed biofi lm for marine heterotrophic bacteria analysis was decimally di- luted, while for coliform and enterococci analysis the samples were fi ltered through a nitro- cellulose fi lter (0.2 μm) and then inoculated onto selective agar plates in duplicates. The number of heterotrophic marine bacteria was determined on Marine agar, DIFCO, USA, (25 °C, 72 h). The number of faecal coliforms was determined on m-FC agar, Biolife, Italy (44.5 °C, 24 h). The number of intestinal enterococci was determined on Slanetz-Bartley agar, Biolife, Italy (35 °C, 72 h). Diatom and phytoplankton analysis For algological analysis, plates were sampled after 1 h, 12 h, 24 h, 48 h, 96 h, one week and one month and placed in sterile bags with care that the biofi lm was not damaged. Foul- ing was scraped off the inner and outer side of the plates using a brush and dispersed in a known volume of 2% formaldehyde solution. Cell counts were obtained by the inverted microscope Zeiss Axiovert 200 using a previously described method (UTERMÖHL 1958). Subsamples of 10 ml were analysed microscopically after sedimentation for 24 h. Micro- plankton (MICRO) cells (longer than 20 μm) were counted under a magnifi cation of 400× (1–2 transects), as well as 200× and 100× (transects along the rest of the counting chamber base plate). The samples were acid cleaned (HCl/H2SO4) prior to qualitative analysis using scanning electron microscope (SEM) Philips 515. For direct plate analyses plexiglass plates were pre-dried at room temperature. Fouling on such prepared plates was observed by SEM Philips 515 and, after being dyed with DAPI, by epifl uorescence microscopy (Zeiss Axio Imager Z1). To get insight into the natural periphytic community, leaves of seagrass Cymodocea nodosa were sampled and phytoplankton samples also, to establish stabile surrounding phytoplankton assemblages. Leaves of seagrass C. nodosa were sampled on the fi rst day of the experiment and the periphytic community was acid cleaned (HCl/H2SO4) and examined by SEM Philips 515. For analysis of the surrounding phytoplankton community, samples were taken at a depth of 5 m by 5 L Niskin bottle after 1 h, 24 h, 48 h, 96 h, one week and one month. Samples of 50 mL were sedimented for 24 h and analysed following the Uter- möhl method (1958) with an inverted microscope (Zeiss Axiovert 200). Pigment analysis In order to gather information about the chemotaxonomic composition of the periphytic community, pigments were analyzed using high performance liquid chromatography (HPLC). Re-suspended samples were fi ltered on 0.7 μm pore size Whatman Glass Fibre Filters (GF/F) and preserved in liquid nitrogen until the analysis. The extraction in 4 mL of cold 90% acetone was performed by sonication, and the extract was clarifi ed by centrifuga- tion. Pigments were separated by reversed phase HPLC following the protocol of BARLOW et al. (1997). Extracts were mixed 1:1 (v/v) with 1 M ammonium acetate and injected into an HPLC system equipped with 3 mm Thermo Hypersil column MOS2 (C-8, 120 A pore size, 150 × 4.6 mm) (Thermo Hypersil-Keystone). Pigments were separated at a fl ow rate of 1 mL min–1 using a linear gradient program with a duration of 40 min. Solvent A consisted of 70:30 (v/v) methanol: 1 M ammonium acetate and solvent B was 100% methanol. Chloro- phyll and carotenoids were detected by absorbance at 440 nm (Spectra System, Model UV MEJDANDŽIĆ M., IVANKOVIĆ T., PFANNKUCHEN M., GODRIJAN J., MARIĆ PFANNKUCHEN D., et al. 412 ACTA BOT. CROAT. 74 (2), 2015 2000). Qualitative and quantitative analyses of individual pigments were performed by ex- ternal standard calibration using authentic pigment standards (VKI, Denmark). Results During sampling, weather conditions varied from day to day, which was refl ected in the structure of the water column and consequently the structure of the periphytic community and biofi lm formation. After 30 days of exposure, a slimy mucous biofi lm was present on plex- iglass plates. The main living constituents of the biofi lm were attached cells of bacteria, cy- anobacteria, diatoms, dinofl agellates, green (Ulvophyceae) and brown (Phaeophyceae) algae. Bacteria Heterotrophic marine bacteria reached the number of 1.96 ± 0.79 × 104 CFU cm–2 dur- ing one month of exposure (Fig. 2A). Despite the fact that faecal coliforms and intestinal enterococci were present in the water column at the beginning of the experiment (faecal Fig. 2. Abundance (expressed as logarithmic value) of heterotrophic marine bacteria through expo- sure time from 1 h to one month (a); Abundance (expressed as logarithmic value) of peri- phytic diatoms through exposition time from 12 h to one month (b); Concentration of pig- ments fucoxanthin and Chl a through exposure time from 24 h to one month (c). COLONIZATION OF DIATOMS AND BACTERIA IN THE NE ADRIATIC ACTA BOT. CROAT. 74 (2), 2015 413 coliforms 255 ± 78 CFU L–1, intestinal enterococci 180 ± 28 CFU L–1) no coliforms were detected in the biofi lm during the entire experiment. Enterococci were present after one month of exposure in a negligible number (60 ± 10 CFU cm–2). Diatoms Periphytic diatoms also showed a steady increase in abundance during the month of exposure of the plexiglass plates (Fig. 2B). The absolute maximum value of abundance of periphytic diatoms was 6.10 × 105 cells cm–2 at 30 days of exposure. The pioneer species Cylindroteca closterium and Nitzschia longissima were determined and they re-occurred with planktonic species Dactyliosolen fragilissimus, Diploneis bom- bus, Proboscia alata and group Pseudo-nitzschia pseudodelicatissima »sensu lato« (Tab. 2). These were followed by Thalassionema nitzschioides, Leptocylindrus danicus, Microta- bella interrupta, Pleurosigma angulatum, Licmophora sp. and Melosira nummuloides. The species with the greatest abundance was Cylindroteca closterium (5.5 × 104 cells cm–2). A total of 30 diatom taxa were determined (Tab. 3) in the periphyton assemblage from the plexiglass plates and the dominant taxa were: Navicula sp. (71%), Thalassiosira sp. (71%), Amphora ovalis (71%), Amphora sp. (43%), Licmophora sp. (43%), Mastogloia sp. (43%), Proschkinia bulnheimii (43%), Thalassionema nitzschioides (43%), Cyclotella sp. (43%). The qualitative analysis of stable diatom community on Cymodocea nodosa re- vealed a total of 10 species of which Amphora ovalis, Cyclophora sp., Navicula sp. and Proschkinia bulnheimii were most frequent. Three taxa were recorded only by analyzing leaves of Cymodocea nodosa: Achnanthes sp., Cyclophora sp. and Microtabella interrupta. Tab. 2. List of all diatom taxa present in biofi lm at different exposure times during research period as well as their abundance. Taxa / Exposure time Abundance (cells cm–2) 1 h 12 h 24 h 48 h 96 h 1 week 1 month Dactyliosolen fragilissimus (Bergon) Hasle 2 Cylindroteca closterium (Ehrenberg) Reimann & J. C. Lewin 16 61 114 151 7234 55764 Diploneis bombus Ehrenberg 2 Leptocylindrus danicus Cleve 26 Licmophora sp. 1426 37176 Melosira nummuloides C. Agardh 703 Microtabella interrupta (Ehrenberg) Round 201 88 Nitzschia longissima (Brébisson) Ralfs 16 25 23 Pleurosigma angulatum (Quekett) W. Smith 2 Proboscia alata (Brightwell) Sundström 65 23 25 Pseudo-nitzschia pseudodelicatissima »sensu lato« 194 163 137 182 Thalassionema nitzschioides (Grunow) Mereschkowsky 25 82 524 780 251 Unindentifi ed pennate diatoms 307 53 225 2234 4909 25722 515814 MEJDANDŽIĆ M., IVANKOVIĆ T., PFANNKUCHEN M., GODRIJAN J., MARIĆ PFANNKUCHEN D., et al. 414 ACTA BOT. CROAT. 74 (2), 2015 Tab. 3. List of all diatom taxa analyzed with SEM at different exposure times, C – diatom taxa found on leaves of seagrass Cymodocea nodosa, Fr. – occurrence frequency of taxa (diatom taxa found on C. nodosa were excluded), 0* – zero frequency for taxa found only on leaves of C. nodosa. Taxa (30 in total) Exposure time C Fr. (%)1 h 12 h 24 h 48 h 96 h 1 week 1 month Achnanthes sp. + 0* Amphora coffeaeformis Cleve + + + + + 58 Amphora ovalis (Kützing) Kützing + + + + + + 71 Amphora sp. + + + + 43 Bacteriastrum sp. + 14 Cocconeis sp. + + + 29 Cyclophora sp. + 0* Cyclophora tenuis Castracane + 14 Cyclotella sp. + + + 43 Cylindrotheca closterium (Ehrenberg) Reimann & J. C. Lewin + + 29 Diploneis bombus Ehrenberg + 14 Diploneis sp. + + + 43 Fallacia sp. + 14 Gomphonema sp. + 14 Grammatophora marina (Lyngbye) Kützing + 14 Licmophora fl abellata (Grev.) C. Agardh + 14 Licmophora sp. + + + 43 Mastogloia sp. + + + + 43 Mastogloia undulata Grunow + 14 Microtabella interrupta (Ehrenberg) Round + 0* Navicula sp. + + + + + 71 Nitzchia sp. + + + 29 Paralia sulcata (Ehrenberg) Cleve + 14 Pleurosigma angulatum (Quekett) W. Smith + 14 Proschkinia bulnheimii (Grunow) Karayeva + + + + 43 Psammodictyon mediterraneum (Hustedt) D. G. Mann + + + 29 Rhizosolenia sp. + 14 Thalassionema nitzschioides (Grunow) Mereschkowsky + + + 43 Thalassiosira sp. + + + + + 71 Triceratium sp. + 14 COLONIZATION OF DIATOMS AND BACTERIA IN THE NE ADRIATIC ACTA BOT. CROAT. 74 (2), 2015 415 The surrounding phytoplankton assemblage was composed of 24 dominant taxa of Bac- illariophyceae, six dominant taxa in the group Dinophyceae and the groups Cryptophyceae and Chlorophyceae. The composition of phytoplankton throughout the exposure period was relatively consistent with the dominant group Pseudo-nitzschia pseudodelicatissima »sensu lato« whose greatest abundance was 10.2 × 104 cells L–1. Biofi lm pigments HPLC pigment analysis revealed 10 different pigments belonging to different groups: peridinin, diadinoxanthin, butinin, chlorophyll a (Chl a), chlorophyll c (Chl c), fucoxan- thin, hexanoiloxifucoxanthin, zeaxanthin, β-carotene and prasinoxanthin, fucoxanthin (con- centration up to 110 ng cm–2) and Chl a (concentration up to 135 ng cm–2) being the domi- nant pigments (Fig. 2C). During this study Chl c, prasinoxanthin and zeaxanthin were detected after 96 h of exposure. Chl a, fucoxanthin and peridinin were dominant from be- ginning. Also, the concentration of fucoxanthin was signifi cantly lower than the concentra- tion of Chl a in the fi rst four days of exposure, and then increasingly reached similar con- centrations (fucoxanthin 108.272 ng cm–2, Chl a 132.424 ng cm–2) (Fig. 2C). Direct biofi lm visualisation During the colonization of the plexiglass plates and biofi lm generation some parts of the surface remained uninhabited, while the others developed a comprehensive three-dimen- sional biofi lm (Figs. 3, 4). Biofi lm thickness was not homogeneous in all parts of the sub- strate and varied from a few μm to a few cm depending on the state of succession on the plates. Discussion Research of periphytic algae on natural and artifi cial substrates is faced with an array of challenges: quantifying the substrate surface as well as the diversity and inaccuracies of the techniques of removing fouling makes it diffi cult to compare published results (BARBIERO 2000). The best practice for studying processes of migration, colonization and growth is: (i) carefully to choose the artifi cial substrate, its material surface, texture and size (CATTANEO and AMIREAULT 1992), (ii) to adjust for easy handling (WEITZEL et al. 1979), (iii) to deter- mine the appropriate location and duration of exposure of the substrate (iv) and to adjust material for the selective attachment of benthic fl ora (SNOEIJS 1991). In our study plexiglass was selected as artifi cial substrate which facilitated removal and the measurement of sur- face fouling, allowed easier handling and reduced the possibility of settling of organisms that are not of interest for the development of primary biofi lms. Moreover, we confi rmed that the plexiglass surface in a marine environment is susceptible to biofouling within 30 days of contact. However, the dominating factor involved in the initial attachment of a bac- terial cell to a surface has remained elusive.Today it is thought that a multitude of factors are involved in processes of settling, including surface conditioning, mass transport, sur- face charge, hydrophobicity, surface roughness and surface micro-topography (PALMER et al. 2007). Our fi nding that faecal coliforms and intestinal enterococci were detected in the water column, but not in a biofi lm may be explained with chemical composition, surface roughness and micro-topography. Coliforms are known to inhabit mixed-population bio- MEJDANDŽIĆ M., IVANKOVIĆ T., PFANNKUCHEN M., GODRIJAN J., MARIĆ PFANNKUCHEN D., et al. 416 ACTA BOT. CROAT. 74 (2), 2015 fi lms in water distribution systems (CAMPER et al. 1996). In laboratory conditions the per- sistence of coliforms in mixed biofi lms on polycarbonate surfaces was highly dependent on the growth rate of the inoculum and type of substratum. Signifi cantly higher numbers of both heterotrophs and coliforms were found on steel (reactive surface) than on polycar- bonate (inert surface) (CAMPER et al. 1996). Bacteria (including coliforms) that initially suc- cessfully colonized the surface can readily acclimatize to low-nutrient conditions (seawa- ter) (NOVITSY and MORITA 1978, KURATH and MORITA 1983, CAMPER et al. 1996). Thus, the absence of faecal coliforms and intestinal enterococci in the biofi lm in our experiment was likely due to the choice of substratum; the mentioned species were unable to initially attach to inert surface of plexiglass and continue their growth in the biofi lm. An important factor during the fi rst week of succession is the strength and direction of the wind because it contributes to the mixing of the water column. The occurrence of plank- Fig. 3. Direct plate analysis under epifl uorescence microscope through the exposure time from 1 h to 1 month; b – bacterial cells; cb – cyanobacterial chains; cc – cells of Cylindrotheca closte- rium; ma – macro-aggregate. Bars = 10 μm. COLONIZATION OF DIATOMS AND BACTERIA IN THE NE ADRIATIC ACTA BOT. CROAT. 74 (2), 2015 417 Fig. 4. Direct plate analysis under scanning electron microscope through the exposure time from 1 h to 1 month; b – bacterial cells; bc – cells of Bacteriastrum sp.; cc – cells of Cylindrotheca closterium; EPS – extracellular polymer substances produced by bacterial and diatom cells; pb – cells of Proschkinia bulnheimii; pd – unidentifi ed pennate diatoms; pl – cells of Pleuro- sigma sp. MEJDANDŽIĆ M., IVANKOVIĆ T., PFANNKUCHEN M., GODRIJAN J., MARIĆ PFANNKUCHEN D., et al. 418 ACTA BOT. CROAT. 74 (2), 2015 tonic diatoms among the pioneer species on the plates might be explained by these environ- mental conditions. In this study we recorded Dactyliosolen fragilissimus, Proboscia alata, Thalassionema nitzschioides, and Leptocylindrus danicus, all common planktonic diatoms of the north-eastern Adriatic Sea (GODRIJAN et al. 2013). Records of frustules of planktonic taxa found in a dense biofi lm can be explained by their natural settling and entanglement after death. Frustules were caught and retained in the dense organic network of the biofi lm. The composition of phytoplankton throughout the exposure period was relatively con- sistent with the dominant group being Pseudo-nitzschia pseudodelicatissima »sensu lato« whose greatest abundance was observed after a month of exposure (10.2 × 104 cells L–1). The group P. pseudodelicatissima »sensu lato« is positively correlated with temperature in the northern Adriatic, as was the case in the western Mediterranean, where higher abun- dances were recorded in the warmer part of the year, for example, in the late summer and autumn (LJUBEŠIĆ et al. 2011, MARIĆ et al. 2012). As succession progressed, planktonic species were replaced by benthic diatoms the abundance of which increased exponentially. The taxonomic composition and abundance of periphytic diatoms in this study confi rms previous research done in the northern Adriatic (TOTTI et al. 2007, CAPUT et al. 2008). According to DENICOLA and MCINTYRE (1990), dia- toms of the genus Licmophora, Cocconeis and Achnantes are among the most prominent colonizers after the fi rst week of exposure. They are accompanied by other motile pennate diatoms which are competitors for light and nutrients such as those from the genus Ampho- ra. As confi rmed by a quantitative analysis of the biofi lm, the qualitative analysis shows that in the initial formation of biofi lms, from the fi rst to the fourth day of immigration, the most important factor is planktonic species (STEVENSON 1986), while later (from the 7th to the 30th day) there is a signifi cant development of benthic diatoms, which are attached with the entire surface of the valve (e. g. genera Cocconeis and Amphora) and EPS production in the form of stands (Licmophora sp.), apical plate, mucilage plates and cell membranes (HO- AGLAND et al. 1993). Additionally MUNDA (2005) in her study near Piran (Gulf of Trieste) describes primary colonizers as Achnanthes and Licmophora species which form dense epi- lithic populations due to their attachment by stalks. Also noteworthy is the centric diatom Paralia sulcata, which is regarded as an indicator species of coastal upwelling situations and is common in plankton and the benthos under low light conditions, which are an advan- tage to it (MUNDA 2005). In our study P. sulcata was recorded at fi rst sampling event when the water column was mostly under wind pressure resulting in mixed conditions. Our study correlates with previous fi ndings (CAPUT et al. 2008) and the abundances of Amphora cof- feaeformis and Navicula veneta that dominated (f ≥ 83%, abundance > 106 cells cm–2) in the Zrmanja estuary. In our study, in addition to these two species, Nitzschia longissima (f = 80%, abundance 1.5 × 106 cells cm–2) was co-dominant. The genera Cyclophora, Achnanthes and species Microtabella interrupta were not re- corded in the biofi lm on the plates at any sampling event. This could be explained by hy- pothesizing that they settle later during the biofi lm formation or rather that they prefer natu- ral substrates such as seagrass. According to ROMAGNOLI et al. (2007) who compared microalgae grown on artifi cial substrata and those that settled on E. racemosum, on average diatom abundance values were signifi cantly lower on artifi cial substrata (2.521 ± 1.784 cells mm–2) than on the hydroid (4.737 ± 3.919 cells mm–2, p < 0.01). The same pattern was observed for diatom biomass, which showed signifi cantly lower values on mimic substrata COLONIZATION OF DIATOMS AND BACTERIA IN THE NE ADRIATIC ACTA BOT. CROAT. 74 (2), 2015 419 (0.033 ± 0.027 μg C mm–2) than on E. racemosum (0.093 ± 0.088 μg C mm–2, p < 0.005). Many studies have shown that macrophytes generally serve as a source of nutrients for periphyton, which makes plexiglass a relatively poor substrate (CATTANEO and KALFF 1979, CARIGNAN and KALFF 1982, BURKHOLDER and WETZEL 1990, CATTANEO and AMIREAULT 1992). The chemotaxonomic analysis of the biofi lm provided an excellent confi rmation meth- od for our microscopic results. Chl c, prasinoxanthin and zeaxanthin were detected after 96 h of exposition which nicely correlated with the appearance of cyanobacteria on the plates after the fi rst week. The presence of Chl a and fucoxanthin from the beginning revealed di- atoms as primary colonizers. Descriptive methods of direct plate visualisation with epifl uorescence and electron mi- croscopy in this study gave a better insight into the sequence of colonization on artifi cial substrates and into the three-dimensional structure of this early biofouling stages. So far, these methods have not been used in research and the data obtained in this study can serve as starting point for the development of such methods in the future. During the fi rst hours of contact with seawater we could observe the deposition of a thin fi lm on the substrate. Sub- sequently, we confi rmed that the bacteria are the fi rst colonizers in the process of immobili- zation (reversible binding), where they fi rst examine the substrate and check for the availa- bility of nutrients, which is followed by the process of consolidation (irreversible binding) when they begin to exude EPS (LEHAITRE and COMPÈRE 2007). On this sticky mixture of proteins, proteoglycans and carbohydrate diatoms start to attach, and they tend to form per- manently adhesive complex structures such as pads, stalks, capsules and tubes. Even though we observed a distinct succession pattern in the biofi lm formation, the rather two-dimen- sional structure of the observed biofi lm seems to indicate, that colonization is a rather indi- vidual process of direct interaction between the cells and the substrate. However the matu- ration of the supposedly organic but abiotic thin fi lm covering the substrate surface might explain the observed succession pattern. Also, differences in a presumed time lag of growth induction after attachment to a surface might contribute to the succession of species ob- served during the biofi lm formation. In conclusion, this study confi rmed that plexiglass surfaces in a marine environment are susceptible to biofouling within 30 days of contact. Abundance of bacterial cells and dia- tom cells increased through the investigated period and they interacted, creating an opti- mized micro-environment. With HPLC analysis this study confi rmed a temporal distribu- tion pattern during the process of biofi lm formation: fi rst colonizers are bacteria and cyanobacteria, and second are diatoms, which together form a primary biofi lm in the sea. In natural biofi lms, niche formation might, in addition to other factors, explain the success and distribution pattern of certain diatoms and associated bacteria. They can support each other by equilibrium of the cross-feeding, possibly optimized by exchange of chemical factors. These associations can be specifi c or random. It is likely that cross-feeding partners may change due to various factors such as light, temperature, water currents etc., or presence of other microorganisms and their secretions. Further, these interactions appear to initialize the formation of diatom biofi lms and aggregates, as has been shown with marine microbial communities. However, we were able to demonstrate a certain spatial independence be- tween bacterial and diatom early settlers, which indicates the importance of abiotic compo- nents in the early biofi lm formation. MEJDANDŽIĆ M., IVANKOVIĆ T., PFANNKUCHEN M., GODRIJAN J., MARIĆ PFANNKUCHEN D., et al. 420 ACTA BOT. CROAT. 74 (2), 2015 Acknowledgments This work was in part supported by Croatian Science Foundation under the project 6433 and under projects of Ministry of Science, Education and Sports of the Republic of Croatia no. 119-1191189-1228, 119-1191155-1203 and 098-0982705-2731. The authors are grate- ful to Paolo Paliaga for his help with epifl uorescence microscopy. References BARBIERO, R. P., 2000: A multi-lake comparison of epilithic diatom communities on natural and artifi cial substrates. Hydrobiologia 438, 157–170. BROTAS, V., PLANTE-CUNY, M-R., 2003: The use of HPLC pigment analysis to study micro- phytobenthos communities. Acta Oecologica 24, S109–S115. BURKHOLDER, J. M., WETZEL, R. G., 1990: Epiphytic alkaline phosphatase activity on natural and artifi cial plants in a P-limited lake: re-evaluation of the role of macrophytes as a phosphorus source for epiphytes. Limnology and Oceanography 3, 736–746. CAMPER, A. K., JONES, W. L., HAYES, Y. T., 1996: Effect of growth conditions and substratum composition on the persistence of coliforms in mixed-population biofi lms. Applied and Environmental Microbiology 62, 4014–4018. CAPUT MIHALIĆ, C., VILIČIĆ, D., AHEL, M., BURIĆ, Z., CARIĆ, M., 2008: Periphytic algae de- velopment in the upper reach of the Zrmanja estuary (Eastern Adriatic Coast), Vie et Milieu – Life and Environment 58, 203–2013. CARIGNAN, R., KALFF, J., 1982: Phosphorus release by submerged macrophytes: Signifi - cance to epiphyton and phytoplankton. Limnology and Oceanography 27, 419–427. CATTANEO, A., AMIREAULT, M. C., 1992: How artifi cial are artifi cial substrata for periphyton? Journal of North American Benthological Society 11, 244–256. CATTANEO, A., KALFF, J., 1979: Primary production of algae growing on natural and artifi cial aquatic plants: a study of interaction between epiphytes and their substrate. Limnology and Oceanography 24, 1031–1037. CHARACKLIS, W. G., COOKSEY, K. E., 1983: Biofi lms and microbial fouling. In: LASKIN A. I. (ed.), Applied microbiology, 93–138. Academic Press. New York. DE NICOLA, D. M. AND MCINTYRE, C. D., 1990: Effects of substrate relief on the distribution in laboratory streams. I. Hydrology. Journal of Phycology 26, 624–633. DE RINCON, O. T., MORRIS, E., DE ROMERO, M., ANDRADE, S., 2001: Effect of ‘pelo de oso’ (Garveia franciscana) on different materials in Lake Maracaibo. NACE International, Corrosion/2001. GODRIJAN, J., MARIĆ, D., TOMAŽIĆ, I., PRECALI, R., PFANNKUCHEN, M., 2013: Seasonal phyto- plankton dynamics in the coastal waters of the north-eastern Adriatic Sea. Journal of Sea Research 77, 32–44 HAMILTON, P. B., DUTHIE, H. C., 1984: Periphyton colonization of rock surfaces in a boreal forest stream studied by scanning electron microscopy and track in autoradiography. Journal of Phycology 20, 525–532. HILLEBRAND, H., SOMMER, U., 1997: Response of epilithic microphytobenthos of the West- ern Baltic Sea to in situ experiments with nutrient enrichment. Marine Ecology Progress Series 160, 35–46. COLONIZATION OF DIATOMS AND BACTERIA IN THE NE ADRIATIC ACTA BOT. CROAT. 74 (2), 2015 421 HOAGLAND, K. D., ROSOWSKI, J. R., GRETZ, M. R., ROEMER, S. C., 1993: Diatom extracellular polymeric substances: function, fi ne structure, chemistry and physiology. Journal of Phycology 29, 537–566. KURATH, G., MORITA, R. Y., 1983: Starvation-survival physiological studies of a marine Pseudomonas sp. Applied and Environmental Microbiology 45, 1206–1211. LEHAITRE, M., COMPÈRE, C., 2007: Biofouling and underwater measurements. In: BABIN, M., ROESTER, C. S., CULLEN, J. J. (eds.), Backscatter 10, 22–25. LEWIS, D. P., PIONTKOWSKI, J. M., STRANEY, R. W, KNOWLTON, J. J., 1997: Use of potassium for treatment and control of zebra mussel infestation in industrial fi re protection water systems. Fire Technology 33, 356–71. LJUBEŠIĆ, Z., BOSAK, S., VILIČIĆ, D., BOROJEVIĆ- KRALJ, K., MARIĆ, D., GODRIJAN, J., UJEVIĆ, I., PEHAREC, P., DJAKOVAC, T., 2011: Ecology and taxonomy of potentially toxic Pseudo- nitzschia species in Lim Bay (north-eastern Adriatic Sea). Harmful Algae 10, 713–722. MARIĆ, D., KRAUS, R., GODRIJAN, J., SUPIĆ, N., DJAKOVAC, T., PRECALI, R., 2012: Phytoplank- ton response to climate and anthropogenic infl uences in the north-eastern Adriatic dur- ing the last four decades. Estuarine, Coastal and Shelf Science 115, 98–112. MUNDA, I. M., 2005: Seasonal fouling by diatoms on artifi cial substrata at different depths near Piran (Gulf of Trieste, Northern Adriatic). Acta Adriatica 46, 137–157. NOVITSY, J. A., MORITA, R. Y., 1978: Possible strategy for the survival of marine bacteria under starvation conditions. Marine Biology 36, 289–295. O’TOOLE, G., KAPLAN, L. A., KOLTER, R., 2000: Biofi lm formation as microbial develop- ment. Annual Review of Microbiology 54, 49–79. PALMER, J., FLINT, S., BROOKS, J., 2007: Bacterial cell attachment, the beginning of a bio- fi lm. Journal of Industrial Microbiology 34, 577–588. ROMAGNOLI, T., BAVESTRELLO, G., CUCCHIARI, E., DE STEFANO, M., DI CAMILLO, C., PENNESI, C., PUCE, S., TOTTI, C., 2007: Microalgal communities epibiontic on the marine hydroid Eudendrium racemosum in the Ligurian Sea, during an annual cycle. Marine Biology 151, 537–552. SABATER, S., GREGORY, S.V., SEDELL, J. R., 1998: Community dynamics and metabolism of benthic algae colonizing wood and rock substrates in a forest stream. Journal of Phycol- ogy 34, 561–567. SLÀDEČKOVÀ, A., 1962: Limnological investigation methods for the periphiton (»auf- wuchs«) community. The Botanical Review 28, 286–350. SNOEIJS, P., 1991: Monitoring pollution effects by diatom community composition: A com- parison of sampling methods. Archiv für Hydrobiologie 121, 497–510. STEVENSON, F. J., 1986: Cycles of soil: carbon, nitrogen, phosphorus, sulphur, micronutri- ents. J. Wiley and Sons. New York. STOODLEY, P., SAUER, K., DAVIES, D. G., COSTERTON, J. W., 2002: Biofi lms as complex dif- ferentiated communities. Annual Review of Microbiology 2002, 187–209. TOTTI, C., CUCCHIARI, E., ROMAGNOLI, T., PENNA, A., 2007: Bloom of Ostreopsis ovata in the Conero Riviera (NW Adriatic Sea). Harmful Algae News 33, 12–13. TOTTI, C., CUCCHIARI, E., DE STEFANO, M., PENNESI, C., ROMAGNOLI, T., BAVESTRELLO, G., 2007: Seasonal variations of epilithic diatoms on different hard substrates, in the northern MEJDANDŽIĆ M., IVANKOVIĆ T., PFANNKUCHEN M., GODRIJAN J., MARIĆ PFANNKUCHEN D., et al. 422 ACTA BOT. CROAT. 74 (2), 2015 Adriatic Sea. Journal of the Marine Biological Association of the United Kingdom 87, 649–658. UTERMÖHL, H., 1958: Zur Vervollkommnung der quantitative Phytoplankton Methodik. Mitteilung Internationa le Vereinigung fuer Theoretische unde Amgewandte Limnolo- gie 9, 1–38. WEITZEL, R. L., SANOCKI, S. L., HOELECEK, H., 1979: Sample replication of periphyton collect- ed from artifi cial substrates, 90–115. In: WEITZEL, R. L. (ed.), Methods and measurements of periphyton communities: A Review, American Society for Testing and Materials.