Currently, it is broadly accepted that plastic pollution impacts the whole marine food web, from zooplankton to marine mammals (Andrady 2011, Cole et al. 2013). Ingestion, entanglement and suffocation of marine wildlife are the best-known problems caused by this type of pollution and are affecting marine habitats worldwide (see reviews by Derraik 2002, Gregory 2009, Cole et al. 2011). Most plastics are resistant to biodegradation, but will break down gradually due to exposure to UV radiation and heat, followed by mechanical action (Billingham et al. 2002, Andrady 2011), simulating components of the plankton community. This gradual fragmentation of plastics causes the release of toxins of the family biophenols or polychlorinated biphenyls (PCBs), which are incorporated into the marine food web (Barnes et al. 2009). PCBs and dichlorodiphenyldichloroethylene (DDE) are also adsorbed to polypropylene (PP) resin pellets from seawater. Plastic pellets act as pollutant accumulators that can transfer hydrophobic pollutants to organisms (Mato et al. 2001, Andrady 2011). A recent study estimates that more than five trillion plastic pieces are floating in the oceans (Eriksen et al. 2014) and this is only the visible face. It is believed that the majority of plastics (70%) end up on the seabed (Hammer et al. 2012).

Marine biofouling is described as the undesired growth of marine organisms on submerged surfaces of anthropogenic origin (Callow and Callow 2002). Surfaces introduced into the marine environment are promptly covered by extracellular polymeric substances (EPS) produced by archaea, bacteria and eukaryotic microbes (Flemming et al. 2007). Polysaccharides, proteins, glycoproteins, glycolipids and other EPS substances create a matrix and retain particulate substances from the environment, providing nutrients for biofilm organisms (Flemming 2009). Micro- and macro-organisms such as diatoms, dinoflagellates, rotifers, arthropods, bivalves, gastropods, worms, tunicates, hydroids, sponges and macroalgae constitute a part of the fouling community. The development of this fouling community on tubes, maritime structures or boats has been causing problems for centuries. Due to the known adverse effects, the species, the biofouling process and also its potential solution have been largely investigated (Callow and Callow 2002, Schultz 2007, Callow and Callow 2011). However, biofouling on marine plastic debris (MPD) has only recently received attention, and current studies (Fortuño et al. 2010, Bravo et al. 2011, Goldstein et al. 2014) highlight the need for more research on the composition of biofouling communities, the structure of MPD and its impact on the marine environment.

To date, two main impacts have been associated with biofouling on MPD: species introduction and community changes. Benthic MPD provides a new colonization substrate for organisms, so MPD has the potential to change the structure of benthic communities and benthic biodiversity in general (Katsanevakis et al. 2007, Sánchez et al. 2013). Floating MPD also provides new habitats, but especially it offers a long-lasting material that can substantially raise the dispersion capacity of organisms (Aliani and Moldcard 2003, Thiel and Gutow 2005, Barnes et al. 2009). The potential large-scale dispersal of MPD, carrying a community of colonizing organisms, provides a vector for the transport of alien species (Barnes 2002, Thiel et al. 2003, Goldstein et al. 2014). Additionally, floating MPD has been suggested as a potential transport vector of harmful algae bloom (HAB) species (Masó et al. 2003).

Recently, the study of the microfouling community has gained attention due to its possible consequences for plastic degradation. The analysis of the microfouling community could provide clues for understanding the sources and fate of MPD. When plastics reach the sea, they are exposed to ultraviolet (UV) radiation and physical weathering leading to their fragmentation into small pieces: what is known as microplastics (Cole et al. 2011). Microfouling can act as a protection against UV radiation, which retards plastic fragmentation (O’Brine and Thompson 2010). On the other hand, microbes can play a role in increasing plastic degradation (Webb et al. 2009, Harshvardhan and Jha 2013).

Among the various processes that affect MPD when it reaches the marine environment, the most influential mechanism is biofouling causing density changes. Density increases with the biofilm development and the subsequent attachment of fouling organisms (Morét-Ferguson et al. 2010). Very little information is available on the causes and consequences of biofilm development. One of the best tools for studying microorganisms in the fouling community is scanning electron microscopy (SEM) (Delgado and Fortuño 1991, Cros and Fortuño 2002).

Three recent papers address for the first time the characterization by SEM of the microfouling community on floating microplastics collected in three different areas: the North Atlantic (Zettler et al. 2013), water around Australia (Reisser et al. 2014) and the North Pacific Gyre (Carson et al. 2013).

The objective of the present work is to further SEM studies by analysing both floating and benthic MPD and to contribute new scientific information on biofouling microorganisms. Detailed analysis of these organisms could provide new insights into the process of colonization and show whether this process continues when pelagic plastics sink to bottom, due to biofouling or due to the weight of the polymer itself, and new organisms are encrusted on the plastic substrate. The colonization of plastics probably also contributes to the modification of benthic ecosystem functioning. Therefore, the potential role of MPD as a new substrate to be colonized by marine organisms has gained importance in view of future effects on biodiversity changes and is also addressed in this study.

The study focuses on four coastal areas in the Mediterranean Sea (Fig. 1), where plastics were sampled during the COMSOM experimental cruise (Sánchez et al. 2013). Results from the previous work by Sánchez et al. (2013) showed high densities of plastics in each area surveyed (i.e. 60 pd/ha on the Catalan coast, northwestern Mediterranean, 59.5 pd/ha on the Central Tyrrhenian coast, 34 pd/ha on the Murcia coast, southern Spanish Mediterranean, and 23 pd/ha on the eastern Ionian coast), representing higher values than those recorded in other areas of the Mediterranean, such as the Adriatic Sea, the Gulf of Lions and eastern Corsica (Galgani et al. 2000). The aim of the present paper is to characterize the microfouling communities of MPD by SEM observations from different marine habitats and to compare for the first time both pelagic and benthic biofouling organisms. The study focuses on the colonization of marine plastics in order to understand their significance as a source of impact on marine ecosystems and their wider environmental implications.

The study is focuses on four coastal areas of the northern Mediterranean, located between 40 and 90 m depth: one in Italy, on the Central Tyrrhenian coast, one in Greece, on the eastern Ionian coast, and two in Spain, on the Murcia and Catalan coasts (Fig. 1) (see further details of the study areas in Demestre et al. 2010, de Juan and Demestre 2012). Samples of surface and benthic macroplastic debris were collected during an oceanographic cruise undertaken from the 21 May to the 24 June 2009.

Samples of macroplastics accumulated over the seabed (hereafter benthic plastics) were collected with an experimental dredge similar to a 2-m beam-trawl, with a 2×40-cm iron-framed aperture and a 10-mm cod-end. To ensure continuous contact of the gear with the seabed, a Scanmar sensor was placed on the iron frame of the dredge. In each study area, 18 dredge samples were randomly collected performing a 15-minute tow at 2.5 kn following the sampling strategy used in epibenthic studies (de Juan et al. 2011). In a previous analysis conducted on these samples, Sánchez et al. (2013) showed that only 8% of collected plastics showed no sign of colonization, while the rest were covered by fouling organisms. A subset of 16 benthic plastics, in the first phase of succession of colonization, was selected to be analysed.

Samples of floating macroplastics (hereafter pelagic) were collected to characterize surface fouling organisms and compare this community with the community of organisms identified on benthic plastics. Surface sampling was performed using an inflatable boat and a hand-net with a 0.5×0.5-m iron-frame aperture and a 1-cm mesh size to collect macroplastics. Sampling followed parallel transects of 20 m length. Each transect was covered for 15 minutes and the overall sampling area overlapped with the area where the benthic samples were collected. A total of 26 pelagic plastics were selected to be studied under the SEM.

For SEM studies, a subsample of the selected benthic and pelagic macroplastics was taken. The dimensions of the plastics were between 225 and 625 mm². Samples were preserved either in 5% buffered formalin or in 2.5% glutaraldehyde buffered in filtered seawater. Samples were dehydrated with increasing concentrations of ethanol, critical-point-dried with CO₂, mounted on aluminium stubs and sputter-coated with gold-palladium. The samples were observed with a Hitachi S-3500N scanning electron microscope operated at 5 kV at the Electron Microscopy Service of the Institute of Marine Sciences (CSIC), in Barcelona (Spain).

The analysis of pelagic and benthic plastics using a scanning electron microscope evidenced that the frequencies of eight large groups of colonizers qualitatively differentiated surface and seabed MPD microfouling communities (Table 1). Diatoms (Fig. 2) appeared in almost 100% of both benthic and pelagic MPD sampled. Dinoflagellates (Fig. 3) occurred in more than 50% of the pelagic MPD sampled, but rarely (13%) on benthic MPD. Coccolithophores (Fig. 4) were attached on both benthic and pelagic MPD, occurring with a relatively high frequency (Table 1). Fungi were the second group in frequency of appearance (Fig. 5L) on pelagic MPD, but this group only appeared sporadically on benthic MPD. Faecal pellets were detected in both domains although they were more frequently recorded on pelagic MPD.

Benthic and pelagic MPD had a very distinct appearance. Pelagic MPD was characterized, in most cases, by a well-developed biofilm containing a high amount of bacteria, fungi and diatoms (Fig. 5M). Bacteria were present in all analysed samples from both pelagic (Fig. 5A, B, C) and benthic MPD. However, benthic MPD was covered by sediment, basically clay (Fig. 6A, C), which complicated identification of organisms. Benthic MPD was characterized by high quantities of sessile and pedunculated protozoans (Fig. 6A, B, C, D), as well as bryozoan colonies (Fig. 6H) (Table 1). Polychaetes (Fig. 6G) were also identified with a moderate frequency and different species of hydrozoans were detected sporadically in both pelagic (Fig. 5D, E, F) and benthic MPD (Fig. 6E). Foraminifera were also identified in several communities (Fig. 6I) from benthic MPD.

The diatoms Ceratoneis, Cocconeis, Navicula, Thalassionema, Achnanthes, Amphora and Licmophora were the genera that characterized the microfouling community of MPD (Table 2, Fig. 2). The most conspicuous difference between the diatom fouling community from pelagic and from benthic plastics was the high occurrence of the pennate diatom Ceratoneis closterium in pelagic samples (73%). The second genera in frequency of appearance on pelagic MPD, Cocconeis were the most frequently detected genera on benthic MPD. Achnanthes and Licmophora, both pedunculate diatoms, were present on ca. 20% of pelagic MPD but they were not detected on benthic MPD. However, a high number of diatoms could not be identified due to their poor state of preservation or because we could not observe their frustule details. The centric genus Cyclotella sp. was detected only in very few cases (Fig. 2H) and Pleurosigma sp. and Entomoneis sp were identified only on one occasion (Fig. 2). Additionally, resting spores were detected in several pelagic samples (Fig. 5) and diatoms symbiotic of hydrozoans were identified on benthic plastics (Fig. 6F).

Thecate and athecate specimens of dinoflagellates were attached especially on pelagic samples (Table 1). Among the thecate dinoflagellates, Prorocentrum minimum (Fig. 3E), Prorocentrum micans, Ceratium sp., Pentapharsodinium tirrenicum (Fig. 3C), Dinophysis sp. (Fig. 3D), Coolia sp. (Fig. 3A, B) and Prorocentrum lima (Fig. 3F) had been described as epiphyte species. Coolia sp. appeared only on pelagic and benthic MPD collected in the Cap of Creus study area. Heterocapsa sp. (Fig. 3G, H) and other unidentified athecate dinoflagellates (Fig. 3I) also appeared in the samples but these organisms were less frequent.

Coccolithophores were identified on both pelagic and benthic MPD (Table 1) but coccospheres appeared only on pelagic MPD (Fig. 4A, B, C, D, G), while individual coccoliths were relatively abundant on benthic MPD. Emiliania huxleyi was the most abundant species on benthic and pelagic MPD, where it appeared in large quantities (Fig. 4K). Several species of Syracosphaera (S. pulchra, S. halldalii, S. molischii) (Fig. 4D, E, F) were highly frequent on pelagic MPD. Additional species identified on pelagic plastics were Syracolithus confusus, Calcidiscus leptoporus, Coronosphaera mediterranea HOL (formerly, Calyptrolithina wettsteinii), Zygosphaera hellenica, Calyptrosphaera dentata, coccoliths of Umbilicosphaera sibogae, and Scyphosphaera apsteinii. Rhabdosphaera clavigera, Helicosphaera carteri HOL (formerly Syracolithus catilliferus) were present on both habitats and Umbellosphaera sp. and Calcidiscus sp. were identified only on benthic MPD. Unidentified holococcoliths were also present.

Other biological structures such as eggs and jellyfish nematocysts were identified attached to pelagic MPD (Fig. 5G, H, I). Pollen grains appeared in some samples in great numbers. Different remains or pieces of copepods or other crustaceans were common organic structures. However, the presence of unidentified organisms or structures was very common (Fig. 5N, O).

In the present study the SEM analysis of the microfouling community on macroplastics across three areas within Mediterranean coastal waters of Greece, Italy and Spain allowed the microfouling community on plastic substrate collected from pelagic and benthic habitats to be compared for the first time. Results indicated substantial differences between benthic and pelagic microfouling communities. We identified six zoological groups colonizing plastics from both habitats, with clear differences in incidence (Table 1). Diatoms were the most abundant biofilm component on MPD from both pelagic and benthic habitats. These organisms are known to maintain and increase the biofilm layer (Bravo et al. 2011, Zettler et al. 2013). Among the 14 genera of diatoms identified, some taxa had been previously described as components of the microfouling community on pelagic MPD (Table 2). Furthermore, most of them have also been described as benthic diatoms: i) members of the epiphytic communities of macrophytes or macroalgae (Majewska et al. 2014), ii) diatoms of the epilithic communities that grow attached to submerged stones (Totti et al. 2007), and iii) diatoms of epipelic communities which reside in the water/sediment interface (Round et al. 1990).

Only six of the identified genera of diatoms were common on MPD from pelagic and benthic domains: Ceratoneis, Cocconeis, Navicula, Achnantes, Amphora and Fragilariopsis. Among them, only the genus Cocconeis showed a relatively high frequency (38%) on benthic plastics. This epiphytic diatom has been described in the Mediterranean as the dominant genus in the epiphytic communities of Posidonia oceanica, The second and third species in abundance on benthic MPD (Amphora and Navicula) are also components of this epiphytic community (Majewska et al. 2014).

The majority of the dinoflagellates detected in our samples (58% in pelagic and 13% in benthic MPD) are also described as epiphytic organisms. Coolia sp., Prorocentrum lima and Prorocentrum micans are rarely found in the water column. However, after turbulent, mixing events, these species can be found in surface waters of Mediterranean beaches (Vila et al. 2001).

Ceratoneis closterium was the most frequently identified diatom on floating MPD. This species is an epipelic diatom residing in the water/sediment interface and rarely found in the water column. This epipelic EPS producer species has been linked with the mucilage events that affect the Adriatic Sea with unusually high frequency (Najdek et al. 2005). Long gelatinous fronts, spreading up to tens of kilometres in length, extend over large areas in the northern Adriatic, causing great economic losses associated with fishing activities and tourism (Giani et al. 2005).

Among the dinoflagellates detected in our samples, a planktonic species, Prorocentrum minimum, is considered harmful due to its capacity to form high biomass blooms at beaches or in harbours with restricted water circulation in the Mediterranean (Vila and Masó 2005).

Our results add new evidence of the role of MPD as potential transport vectors for noxious events in the Mediterranean (Masó et al. 2003). In the Mediterranean, the majority of HAB events are produced by dinoflagellates, whose life-history strategies play a key role in combination with meteorological and marine physical features, such as fronts, convergences (Basterretxea et al. 2005, Masó and Garcés 2006) and sites with low water renewal such as harbours (Vila et al. 2001). The same physical processes that permit microalgae to develop high biomass blooms or mucilage events act on floating plastics, favouring their accumulation and giving microalgae the opportunity to attach to the abiotic substrate.

The nature of the synthetic polymer (Browne et al. 2010), its size, shape and its colonization by biofouling organisms (Morét-Ferguson et al. 2010, Lobelle and Cunliffe 2011) determine the buoyancy of the MPD and, consequently, its surface transport by wind and currents or its sinking to the seafloor. Along its journey through the marine environment, MPD provides a new substrate on which a succession of colonizing organisms can find a suitable habitat to persist, forming what Goldstein et al. (2014) called “islands” of substrate-associated organisms. The process of colonization causes density changes in MPD, which, in combination with the effect of currents, may cause their sinking to the seabed. Once on the seabed, the process of colonization continues (Fortuño et al. 2010, Demestre and Masó 2012, Sánchez et al. 2013). Not only macroplastics but also microplastics are subject to microfouling (Reisser et al., 2014). During the SPURS-MIDAS (March-April 2013) oceanographic cruise in the North Atlantic subtropical gyre (Font et al. 2014), neuston samples were collected in order to capture microplastics and see the degree of colonization by microorganisms. Well-developed microfouling communities were observed on most of the microplastic captured (Masó, pers. comm.).

Sánchez et al. (2013) described changes of the soft-bottom macro-organisms communities due to the accumulation of MPD. The colonization of plastic debris by primarily inhabitants of rocky or gravelly bottoms could have further consequences for the ecosystem functioning, e.g. by modifying the habitat at a small spatial scale and facilitating the introduction of species, leading to new negative/positive species interactions. In the pelagic domain these potential changes are difficult to assess, but our results from SEM analyses of microfouling communities indicate that pelagic MPD was colonized by members of the epilithic, epiphytic and epipelic microalgae communities (Round 1990). In their original environment these microalgae deal with high levels of turbulence, and the secretion of EPS is an adaptive strategy in these habitats (Consalvey et al. 2004). Therefore, it is not surprising that plastic is a suitable habitat for them. MPD collected in the pelagic environment could have been colonized on the seabed in the near-coastal zone and refloated due to meteorological conditions or due to a gain in buoyancy caused by grazing. Floating plastics will be advected to the shoreline by daily breeze conditions or by local meteorological events. When breeze conditions reverse (wind flowing from earth to open sea), they would be washed offshore and dispersed, so the microfouling organisms can extend their distribution range and/or be the inoculum for the next bloom event. Although our results showed that there were some common species on pelagic and benthic MPD, we were unable to elucidate the source of the fouling communities. The process of sinking and refloating could be a key issue in the study of plastic fate, and therefore needs further investigation.

It is widely recognized that plastic pollution is a global problem and, though research on this topic has increased dramatically in recent decades (Derraik 2002, Cole et al. 2011, Cózar et at. 2015), we are still far from understanding its real consequences. Our results show that the microfouling communities attached to both benthic and floating MPD are dominated by benthic diatom genera. How this highly available substrate would modify their biogeography is an open issue. In fact, remarkably little is yet known about geographic distribution of diatoms (Vanormelingen et al. 2008), and in future studies MPD should be considered as a dispersion vector. Overall, our results highlight the need for further research in order to elucidate the role of MPD in increasingly harmful events at a global scale, either by extending the range of distribution of noxious species or by favouring some species rather than others.