Microalgae colonization and trace element accumulation on the plastisphere of marine plastic debris in Monastir Bay (Eastern Tunisia)

In this study, we examined the toxicity potential of the epiplastic microalgal community that developed on low-density polyethylene (LD-PE) plastic debris found in two distinct regions of the Monastir Bay (Tunisia): the coast exposed to anthropogenic discharges and the open sea in front of the Kuriat Islands. Concentrations of potentially toxic elements (PTEs) accumulated in sediments and plastisphere were compared in order to determine their toxicity potential to biological life. The collected plastispheres were predominantly composed of cyanobacteria, chlorophytes, and diatoms. Diatoms display a relatively high diversity (25 species). At all stations, potentially harmful microalgae (PHM) were more abundant in the plastisphere than in seawater and the coastal zone seems to harbour increased number of potentially harmful cyanobacteria within the plastisphere. At the offshore station S1, the PHM community was dominated by the potentially harmful diatoms belonging to the genus Pseudo-nitzschia. Phormidium sp. was the main potentially harmful cyanobacterium identified in the plastisphere of S1. PTEs concentration in the plastisphere was higher than in sediment and ranking with very high contamination factors at all sites according to the sequence Pb > Cu > Cd > Ni > Zn. The highest accumulation of PTEs in the plastisphere was recorded near harbors and industrial zones with important human interference. This work shows that plastisphere can be a threat to vulnerable species not only because it can contain PHM but also because it can accumulate PTEs.


Introduction
The Mediterranean Sea has been described as the sixth gyre area of marine litter accumulation (Cózar et al. 2015) with plastic debris averaging 423 g /km 2 . Indeed, due to surface currents entering into the Mediterranean Sea via the Strait of Gibraltar, while deep currents exit via the Atlantic Ocean (Ben Ismail et al. 2012), plastic debris (23.150 tons) is trapped in the basin and carried by currents in a closed loop (Eriksen et al. 2014).
Currently, between 1000 and 3000 tons of plastic debris float on the surface of the Mediterranean: fragments of bottles, bags, packaging, fishing lines, etc. (Jambeck et al. 2015;Abreu and Pedrotti 2019). According to a study by the World Wide Fund for Nature (WWF), Tunisia is the fourth largest per capita consumer of plastic products in the Mediterranean. In 2016, Tunisia generated 0.25 million tons of plastic waste, of which 0.05 million tons (20%) was uncollected and 0.20 million tons (80%) was collected for treatment. About 0.15 million tons (60%) of this waste was sent to landfills; 0.04 million tons (16%) was openly discarded; and only 0.01 million tons (4%) was recycled. The Tunisian economy loses about $20 million annually due to plastic pollution, which affects tourism, shipping, and fishing.
Moreover, environmentalists had warned that, in the years of this global pandemic, the plastic pollution from masks and gloves could pose a severe threat to the marine environment and life as the pollution mounting rapidly (Amaral-Zettler et al. 2020;Das et al. 2021). For example, biofilms (bacteria, viruses, phytoplankton, micropredators) and marine invertebrates use plastic debris of various sizes (macro-, meso-, and microplastics) as a habitat. These substrates have also been shown to bind pathogens (Zettler et al. 2013;Amaral-Zettler et al. 2020) and chemical, metallic, and organic contaminants that accumulate into the food chain after ingestion by aquatic organisms, eventually causing infections, strangulation, internal injuries, and hormonal disruption (Andrady 2011; Rochman et al. 2013).
Recent studies have confirmed not only the colonization of plastic surfaces by specific communities compared to those in the surrounding seawater (Zettler et al. 2013), but also those in the sediment (Stefatos et al. 1999;Harrison et al. 2014;Reisser et al. 2014;Oberbeckmann et al. 2016;De Tender et al. 2017 andDebroas et al. 2017). They have also reported the presence of invasive species, potentially toxic microalgae (Masó et al. 2003(Masó et al. , 2016, marine insect eggs (Ribeiro-Brasil et al. 2022), and halobates (Zettler et al. 2011). In addition, the biofilms present in the microplastics allow the adsorption of PTEs (Bradney et al. 2019). Several factors can contribute to the adsorption of PTEs, such as the presence of surface charges, the oxidation state of trace elements, and the pH and salinity of seawater (Bradney et al. 2019). Small plastics in which relatively high concentrations of PTEs accumulate can easily be ingested by living organisms (Cole et al. 2015;Lusher et al. 2015;Amélineau et al. 2016;Nelms et al. 2018) and be a vehicle for contamination of the marine trophic chains. However, this is a subject that is still poorly understood.
Although scientific research on the quantity, distribution, degradation, and sequestration processes of plastic debris and characterization of microbial communities that colonize plastic debris (macro-, meso-, and microorganisms) has increased spectacularly in recent decades, studies remain limited to the European continent and some countries in America and Asia, and none from Africa. Some studies have monitored many plastic along the coast and/or open ocean (Ng and Obbard 2006;Zhou et al. 2018;Curren and Leong 2019), but the toxicity potential of the epiplastic microalgal community is still under investigate.
In this work we aimed to (i) compare the composition and structure of the microalgal community attached to the plastisphere along the eastern shore of Monastir Bay (Tunisia, northern coast of Africa and southern area of the Mediterranean Sea) with that found in the surrounding seawater; (ii) compare the capacity of the plastisphere to accumulate potentially toxic elements (PTEs) along with that of the marine sediment. Monastir Bay was chosen because it is a semi-enclosed ecosystem with a weak hydrodynamic regime due to its sub-marine topography (Souissi et al. 2014), which makes it a suitable environment for plastic debris contamination and plastisphere development. In addition, this bay is impacted by pollutants due to anthropic activities (Challouf et al. 2018;Damak et al. 2020).

Study site
Monastir Bay is considered a semi-enclosed area, bounded on the northwest by the Fast Island of Monastir and on the southeast by the shoals connecting the Fast Island of Thapsus with the Kuriat Islands ( Fig. 1).
Monastir Bay is the seat of increasing urbanization. The population density (approximately 600,000 inhabitants in 2018) is one of the highest in the country. Several buildings, municipalities, industries, and sewage treatment plants line the bay coastline, exerting strong environmental pressure on it. Currently, there are 500 companies in the textile sector, including about 35 specialized in the chemical treatment of fabrics (especially jeans), scattered throughout the region. It is also the site of large-scale coastal development and important fishing activities with 4 fishing ports: the fishing ports of Monastir, Ksibet El Mediouni, Sayada, and Teboulba. In addition, aquaculture is well-developed in the area and has a significant environmental impact on the bay (Challouf et al. 2018;Damak et al. 2020). Monastir Bay has become the most important offshore aquaculture area, where all farms raise perch and sea bream.

Sampling
The sampling sites in Monastir Bay were chosen based on the degree of pollution and the nature of the polluting activity. Eight sampling stations (S1-S8) were selected: seven (S2-S8) along the coast and one (S1) in the inner offshore area. The coastal stations are characterized by a muddy bottom and high eutrophication conditions due to anthropogenic influences. Fly dumping at S2 and S7; building construction debris at S3; dumping flies, maintaining fishing boats, and painting at S4; discharge of treated water from the wastewater treatment plant into S5; painting and maintenance of boats, as well as cleaning and maintaining offshore aquaculture cages at S6 and the vicinity of the fishing harbor of Bekalta at S8. The offshore station S1 is relatively balanced and characterized by the presence of offshore aquaculture cages in its eastern part. Three different matrices were selected for sampling, namely coastal seawater, plastisphere, and sediment.
From the coastal seawater, 3 L were collected: 2 L were utilized for chlorophyll a and nutrient analysis; and 1 L was utilized for microalgae enumeration. A 3‰ formaldehyde solution was used to preserve the microalgae samples. For each station, between 3 and 6 white LD-PE (low-density polyethylene # 4) plastic bags floating on the sea surface and also in surface sediment and exhibiting a "mature" plastisphere (green-yellow patches conspicuous to the eye) were collected using a stainless-steel grab. Each plastic bag is cut into two pieces with stainless steel scissors. Two batches were obtained for each sample. The first batch was placed in a glass bottle containing sterile seawater previously filtered onto a 0.22 µm porosity filter. This batch was preserved with a 3‰ formaldehyde solution and used to count microalgae in the plastisphere. The second batch was placed into a glass bottle previously rinsed with HCl solution (1 M) and Milli-Q water. To stop biological reactions, three drops of highpurity nitric acid were added/diluted in the second batch in each bottle. This batch was used for trace element analysis on the plastisphere. Three sediment replicates with cylindrical plastic cores (7 cm in diameter and 10 cm long) were collected at each site. The top 3 cm of the core was carefully cut off and placed in a plastic container for trace element analysis. All samples were stored in the dark at 4 °C before being taken to the laboratory.
During sample collection, all relevant parameters were recorded including coordinates, GPS, and weather conditions. Chemical parameters of coastal water (temperature, pH, salinity, and dissolved oxygen (DO) were recorded in situ at each station using a multiparameter kit (Multi 3400 i/SET; sensitivity (± 1digit) especially important for pH (0.01).

Nutrients and chlorophyll a analysis
Nutrient analyses were performed using the BRAN and the LUEBBE type III Autoanalyzer. 1 L of seawater adjacent to each LDPE plastisphere sample was collected, then filtered through a fiberglass filter (Whatman GF/F) and stored at − 20 °C until analysis. The determination of nutrient concentrations was performed using a colorimetric assay according to the method of Grasshoff et al. (1983). The quantification of total chlorophyll a concentration was performed as described by Lorenzen (1966) and Aminot and Chaussepied (1983). The absorbance of pigment extract was measured at 665 nm and 750 nm both before and after acidification. The equation used was: [Chl a] μg/L = 26.7 × (λ 665 − λ 750) − (λ' 665 − λ' 750) × 10 / V, where λ: absorbance before acidification and λ': absorbance after acidification.

Plastisphere extraction
Different techniques were used to detach the plastisphere from the plastic surface: successive washes with a spray of filtered seawater, ultrasound baths at different intensities, and manual scrapings with glass slides, toothbrushes, or cytological brushes. To evaluate the effectiveness of each technique, observations of the plastic surface before and after were made using optical and electron microscopy. Microscopic observations revealed that brushing with a cytological brush offers the most efficient tool for the purpose of the study (Fig. 2).

FTIR-ATR analysis
An attenuated total reflectance (ATR) and Fourier transform infrared (FTIR), study, was conducted to confirm that the debris from the recovered plastic bags is LDPE. Before being subjected to ATR-FTIR analysis, the samples were cleaned with a clean room towel dampened with deionized water from an LDPE bottle (Jung et al. 2018).
The ATR-FTIR spectra of the plastic debris were examined using a 45° incidence Bruker FTIR spectrometer (Equinox 55, Bruker Co., Ettlingen, Germany). Spectra were recorded between 4000 and 400 cm −1 , with 32 scans total and a resolution of 4 cm −1 . Peak height algorithms were used to identify the absorption bands, which were then recorded and compared to the literature-reported absorption bands for each polymer (Chércoles Asensio et al. 2009;Jung et al. 2018;and Nandiyanto et al. 2019).

SEM-EDX analysis
Scanning electron microscopy with energy-dispersive X-ray was used to observe the microalgal community growing on the plastisphere and to evaluate the effectiveness of plastisphere extraction techniques. To prepare samples for SEM the methodology described by Kirstein et al. (2018) was used with some modifications. Briefly, 0.5 cm 2 strips of LD-PE covered with plastisphere were fixed at 4 °C for 10 days in phosphate buffer solution (0.1 M), pH = 7.2, containing 3% glutaraldehyde. Strips were dehydrated in ethanol baths of increasing concentration (30%, 50%, 70%, and 90%) for 10 min, followed by three 100% baths. Samples were dried under a vacuum using a paddle pump and then observed using a Q 250 Analytical SEM (Thermo Scientific™), which gives qualitative and quantitative information as to elemental composition.

Microalgae: morphotaxonomic characterization and cell enumeration
To recover the epiplastic microalgae from the plastisphere, we used 10 g of LD-PE covered with plastisphere for each sample. The sample was placed into a glass crystallizer, and 200 ml of seawater filtered through a 0.22 µm filter was added. The sample was gently scraped off with a sterile cytologic brush. After detachment, a volume of filtered seawater was added to the plastisphere solution to reach a final volume of 1L. Recovered samples were stored in a 3‰ solution of formaldehyde in darkness and at room temperature for subsequent enumeration.
Morphotaxonomic characterization ensures microalgal enumeration in both matrices, thus allowing the quantification of groups and species along with the calculation of relative abundance to compare the variation between the two matrices and among different stations. We used the classical microalgal identification technique (Uthermohl 1958) to determine the composition of the microalgal community growing on the plastisphere and in seawater. Samples were counted with an inverted microscope (Nikon Eclipse TS100) at a final magnification of 200 x. Cell abundance was expressed per g fresh weight for plastisphere (cells g −1 fw) and per litter (cells L −1 ) of seawater. Microalgal taxonomy was made according to identification keys (Hasle et al.1996;Komárek and Anagnostidis 2005).

Trace element analysis
After being oven dried at 50 °C, the samples were homogenized and sieved through a 2-mm mesh screen. To determine total trace elements, 0.5 g of plastisphere and 0.5 g of sediment samples were digested completely with a mixture of concentrated solutions of HF, HClO 4 , and HNO 3 , as described by Mudroch et al. (1996). The final solution was diluted to 100 mL in volume. The concentrations of Cd, Co, Cu, Cr, Ni, Pb, and Zn were determined using ICP-OES (Perkin Elmer ICP optima 8000). The ICP-OES was calibrated using multi-element calibration standard (Perkin Elmer). The two certified reference materials, BCR-32 (Natural Moroccan Phosphate Rock: Phosphorite) and BCR 414 (plankton) were used to estimate the recovery of the method. The measured recovery rates were between 91.50 and 101.90% for BCR 32 and between 93.07 and 106.29% for BCR 414. Each sample was analysed 3 times to verify the accuracy of the instrument. The relative standard deviation of the three replicates was calculated. For all elements, the RSDs were between 0.16 and 4.9% (Table 1).

Geoaccumulation index
In the coastal area, the use of the geoaccumulation index (Igeo) introduced by Müller and Suess (1979), was calculated using the equation: where C m is the metal concentration and C b is the geochemical background concentration of the metal (Turekian and Wedepohl 1961). The Igeo values are classified into seven categories, each representing a degree of pollution: Igeo < 0: no pollution; 0 < Igeo < 1: no pollution or moderate pollution; 1 < Igeo < 2: moderate pollution; 2 < Igeo < 3: moderate to strong pollution; 3 < Igeo < 4: heavy pollution; 4 < Igeo < 5: strong to extreme pollution; and Igeo > 5: extreme pollution.

Contamination factor
To describe the contamination of toxic trace elements in a marine environment, we may define a contamination factor (C i f ) accordingly: The mean content of the trace elements (C i m for Cd, Pb, Cu, Zn) from the plastisphere was calculated in µg/g; C i n = concentration reference value for those metals was taken from the distribution of the elements in some major units of the earth's crust, background value or I = 2 ( ∕1.5 ).
The classification of the contamination factors was estimated according to the following scale: C i f < 1-low contamination factor; 1 < C i f < 3-moderate contamination factor; 3 < C i f < 6-considerable contamination factor; 6 ≥ C i f -very high contamination factor.

Statistical analysis
The data were logarithmically transformed with log (X + 1) prior to statistical analysis. The data was presented as the mean ± standard deviation. The considered variables were tested using Pearson's coefficient. The results were statistically verified using the STATISTICA 12 software package. For all statistical purposes, the criterion of significance was set at P < 0.05.

Physicochemical and nutriments characteristics
The results of the physicochemical parameters and nutrients in Monastir Bay were presented in Table 2. Temperature, pH, and salinity indicated very close and similar concentrations between all stations. The average temperature was 30.06 ± 2.7 °C, the pH was 8.34 ± 0.79, and salinity ranged from 37.7 ± 3.39 to 38.9 ± 3.89 mg/l. Concerning oxygen concentrations, stations S3, S4, S5, and S6 (extending from Khniss to Lamta) exhibit the lowest values with 14.50 ± 1.16, 13.10 ± 1.31, 13.60 ± 0.95, and 13.60 ± 1.08 mg/l, respectively. The rest of the stations showed values around a : the Pb concentration value indicated in the BCR-032 is an estimated value and not a certified value. b : the Co concentration value indicated in the BCR-414 is an estimated value and not a certified value. DLs detection limits, RSD relative standard deviation 16 mg/l, mainly Kuriat Island (S1) and Ras Dimas (S8), with 16.80 ± 1.68 mg/l. This variability in oxygen concentrations among stations will accordingly influence metals bioavailability within the water-sediment interface (Zhang et al. 2014;Zaaboub et al. 2015) through the formation of Mn oxides and hydroxides, for example. In terms of nutrients stations, S4 (in front of Ksibat El Mediouni) and S5 (in front of Lamta wastewater treatment plant) showed significant differences from the other stations. S5 has the highest concentrations of NO 2 − (4.59 ± 0.4 µmol/l), NO 3 − (15.49 ± 1.54 µmol/l), silicates (10.54 1.05 µmol/l), total nitrogen (39.45 ± 3.94 µmol/l), and total phosphorous (16.04 ± 1.60 µmol/l). When combined with temperature and salinity, nutrients can strongly shape the community attached to the plastic, being generally associated with high nutrient levels (Oberbeckmann et al. 2018;Li et al. 2019). The distribution of chlorophyll revealed that the highest concentration was recorded at station S3 (Khniss) at 6.81 ± 0.68 mg/m 3 . The low oxygen stations (S3, S4, and S5) exhibit the highest concentrations of Chl a and nutrients because they receive large amounts of waste from domestic sewage discharges, which promotes oxygen consumption by bacteria. All these observations underline the importance of eutrophication observed between Khniss (S3) and Ksibat El Mediouni (S4) due to deleterious conditions and extremely high nutrient levels. We consider that the sum of these factors will shape the plateau and determine the rate of metal accumulation (Nava and Leoni 2021).

FTIR-ATR
The infrared spectra and absorption bands of the LDPE plastic bag labelled with resin code 4 (control) and the LDPE plastic bag debris collected from all stations ( Fig. 3) show that the corresponding spectra exhibit a general allure similar to LDPE (Jung et al. 2018). This allure is characterized by asymmetric stretching of CH 2 at 2920 cm −1 , symmetric stretching of CH 2 at 2845 cm −1 , bending deformation of CH 2 at 1465 cm −1 , symmetric deformation of CH 3 at 1377 cm −1 , and rocking deformation of CH 2 at 717 cm −1 .
In particular, the spectra of stations S3 to S8 showed the formation of two new functional groups. The first is represented by a shoulder in the region of 1680-1630 cm −1 . This corresponds to amide, a form of carbonyl group (Nandiyanto et al. 2019). The second is represented by a shoulder in the 1100-1070 cm −1 region and a peak at 825 cm −1 . This corresponds to C-O stretching and C-H stretching, respectively, a form of oxy compound (Coates 2006). The formation of a carbonyl compound and oxy compounds is a sign of oxidation of LDPE. It is noteworthy that the shoulder in the 1100-1070 cm −1 range can be attributed to expanded polystyrene (EPS) secreted by microorganisms. Indeed, FTIR spectra in the 900-1300 cm −1 range are characteristic

SEM-EDX analysis
Scanning electron microscopes (SEM) with an energy dispersive X-ray (EDX) microphotograph highlights the richness and the diversity in the shapes and sizes of prokaryotic and eukaryotic microorganisms living on the plastisphere along the coast of Monastir Bay. For example, we found rodshaped, spherical, and coccoid bacteria, filamentous cyanobacteria, and centric and araphid pennate diatoms (Fig. 4). However, no dinoflagellates were found using SEM, presumably because they were washed away during SEM processing. The fibrous and mucilaginous structures of extracellular polymeric substances secreted by microorganisms can also be seen. Indeed, EPS is a necessary compound for the formation of the plastisphere and can be referred to as "the cement" of this new habitat. EPS ensures cell attachment to their plastic support and nutrient chelation for microorganism growth. Moreover, SEM-EDX is also useful to confirm the effectiveness of the plastisphere extraction technique. Indeed, through brushing the extracted plastisphere ( Fig. 2) will be easily subjected to an assessment of the diversity of the epiplastic microalgal community and the quantification of the accumulated trace elements. For example, the elemental analysis results show that the amount of trace elements in the brushed plastic was zero after plastisphere extraction. SEM has been already used by Zettler et al. (2013) to characterize plastispherecolonizing organisms. Soon after, scientists used SEM to track the progression and kinetics of plastisphere formation (Amaral-zettler et al. 2020;Ramsperger et al. 2020 andCheng et al. 2021). This technique provides a complete description of microorganism diversity and activity during the three stages of biofilm formation on plastics (primo-colonization, growing, and maturation phases). In this sense, our preliminary results of the elemental analysis and the concentration of trace elements determined by ICP-OES showed a good correlation between the mass percent of elements and the concentration of heavy metals (Data not displayed). These results are encouraging to explore the possibility of using elemental analysis to follow the accumulation kinetics of heavy metals on the plastisphere.

Microalgae: community composition, morpho-taxonomic characterization, and cell counts in the two matrices: seawater and plastisphere
Several parameters influence the composition and structure of the community of microorganisms that harbour plastic debris. Among these parameters, the plastic size, the habitat, the physicochemical quality of the water, the type of plastics, the successive phases of maturation of the biofilm, and even the sampling period (De Tender et al. 2015;Oberbeckmann et al. 2016;Hoellein et al. 2017;Frère et al. 2018).
Moreover, different modern techniques are currently used to study the structural complexity of microbial communities growing on the plastisphere. Some scientists have used variety of molecular methods (Clone-libraries, Denaturing Gradient Gel Electrophoresis System-DGGE, Amplicon sequencing, and Shotgun metagenomics). Others used microscopic techniques such as SEM and light microscopy coupled with FISH. Each of these methods has its advantages (good taxonomic, multiple samples analysed simultaneously) and disadvantages (a minor part of microorganism diversity identified, expensive instrumentation, and limit in the taxonomic resolution) (De Tender et al. 2017).
To investigate the composition and the spatial structure of the microalgae community growing on the plastisphere of plastic debris collected in July 2020 in the coastal marine waters of Monastir Bay, we limited our work to white lowdensity polyethylene (LD-PE; floating plastic), and we used the classical techniques of microalgae identification based on microalgal cells counting by a reverse-phase microscope.
Morphotaxonomic characterization ensures the enumeration of microalgal cells in both matrices, which facilitates group and species quantification. It also allows the calculation of relative abundance to compare the variation between the two matrices and between different stations.
To analyse and compare the microalgae community present in the two matrices: plastisphere (P) and seawater (SW), the relative abundance (RA) of the different microalgae groups (cyanobacteria, diatoms, chlorophytes, dinoflagellates, and euglenophyceae) to total microalgae was calculated (Fig. 5a).
The redundancy analysis (RA) shows dissimilarities between the different groups in the two matrices, which confirm the specificity of the community structure living on the plastisphere, compared to seawater, and revealed the spatial variation of the microalgae community according to the sampling stations. In parallel, the morphological-taxonomic characterization revealed the specificity of the group structures formed in the plastisphere compared to seawater.
The matrix sea water (SW) was composed of different groups, i.e., cyanobacteria, diatoms, and dinoflagellates. Euglenophyceae were also observed with a relatively low relative abundance (RA) compared to the other groups, while Chlorophyta was absent. The highest RA for cyanobacteria (91%) was observed at station S2, while the highest RA for diatoms (79.03%) was observed at S1. For dinoflagellates, the highest RA (49.74%) was found at S8.
In the plastisphere, the number of cyanobacterial species varied from 3 to 8, depending on the station. Although the white LD-PE was the support for all the plastispheres investigated in this work, the taxonomic composition was highly variable among stations. There is no clear dominance of one or two species, as was the case for the seawater matrix. Anabaena sp. was the only cyanobacterium observed on all the collected plastispheres with an average abundance of 3.25 × 10 7 cells g −1 . Its RA varied from 1.3% for station S7 to 41.66% for station S4. Some cyanobacteria species are observed only on one plastisphere. Calothrix sp., Closterium sp., Leptolyngbya sp., and Tolypothrix sp. are only observed on the plastisphere of station S3. While, Gloeocystis sp. and Gomphosphaeria sp., are only present on the plastisphere of stations S8 and S6, respectively.
Our analysis also shows that the eight plastispheres collected were hosted by the potentially harmful cyanobacteria: Lungbya majuscula and Phormidium sp. with a clear dominance of Phormidium sp. with an average abundance of 3.24 × 10 7 cells g −1 . Our results are consistent with the earlier observations in marine studies that identified cyanobacteria as one of the most common photoautotrophic prokaryotes found on marine plastic debris (Zettler et al. 2013;Oberbeckmann et al. 2014;Bryant et al. 2016;Oberbeckmann et al. 2016;De Tender et al. 2017;Dussud et al. 2018a;Muthukrishnan et al. 2019). Yokota et al. (2017) explained the important role of cyanobacteria in ensuring the buoyancy of plastic debris. Indeed, cyanobacteria endowed with gas vacuoles can temporarily increase the overall buoyancy of plastic debris until they die or are displaced or consumed by higher density organisms.
The two other groups namely, chlorophytes and diatoms were also found on the plastisphere matrix. The occurrence of Euglenophyceae was also recorded with a relatively low RA. The highest RA for Euglenophyceae (11.98%) was registered for station S4. Chlorophyceae were recorded from S1 to S5. The highest RA for Chlorophyta (89.27%) was noted for station S1 near the Kuriat Islands. The two Chlorophyceae found on are two non-mobile green microalgae found in freshwater environments, namely Scenedesmus sp. and Pediastrum sp. Their abundance averaged 2.16 × 10 5 and 1.3 × 10 5 cells g −1 , respectively. Chlorophyceae have rarely been observed on the plastisphere, with a few data reported in both freshwater and seawater plastispheres (Zettler et al. 2013;Kettner et al. 2017;Weig et al. 2021).
Our findings show that dinoflagellates are almost absent (an RA ≤ 0.90%) in all the sampling plastispheres. Regardless of the sampling stations, the analysis of the relative abundance of dinoflagellate species (Fig. 5c) demonstrated that the seawater had a dinoflagellate "mosaic" of 26 species, whereas the plastisphere was characterized by low species diversity (only 4 species and unknown Dinocystis). As can be seen, the plastispheres collected at stations S2, S4, and S8 do not contain dinoflagellate species. Although dinoflagellates made up a larger portion of the community throughout the water column, these taxa became more common in the plastispheres during the course of the time series (Masó et al. 2003;Zettler et al. 2013;Reisser et al. 2014;Kettner et al. 2017;Dudek et al. 2020). However, not all dinoflagellates were related to both plastic-attached and water column communities.
The taxonomic composition of the dinoflagellates was dissimilar to that of the plastisphere and the surrounding seawater. For example, for station S1, the seawater matrix is composed of 9 species of dinoflagellates: Alexandrium minutum, Alexandrium pseudogonyaulax, Gymnodinium catenatum, Peridinium sp., Polykrikos sp., Prorocentrum cordatum, Prorocentrum rathymum, Protoperidinium minutum, and Protoperidinium pyriforme, and a few hundred dinocysts with an RA = 5%, while the plastisphere matrix contains only Indeterminate dinocysts. Resting cystis of unidentified dinoflagellates and both temporary cysts and vegetative cells of Alexandrium taylori were enumerated on floating plastic debris collected in summer during a bloom of harmful algae (HAB) Alexandrium taylori along the Catalan coast (north-western Mediterranean) (Masó et al. 2003). The production of temporary cysts has typically been linked to unfavourable environmental factors. This enables the population to endure the worst possible weather circumstances and survive. According to Garcés et al. (2002), stickiness may also be important for the dispersal of phytoplankton species that have an adhesive stage in their life cycle.
The planktonic species Alexandrium minutum and the epibenthic species Coolia monotis, Prorocentrum concavum and Prorocentrum lima are the only dinoflagellates species found on the plastisphere with an average abundance of 6.00 × 10 2 , 1.2 × 10 3 , 1.5 × 10 2 , and 1.8 × 10 3 cells g −1 , respectively. All dinoflagellate species found on the collected plastisphere are autotrophic. However, Dudek et al. (2020) demonstrated that the plastisphere communities observed in the Caribbean coastal site were dominated by the heterotrophic dinoflagellate Amphidinium sp. at weeks 1 and 3, but shifted to the autotrophic strains Prorocentrum sp. and Alexandrium sp. at week 6 of incubation. This means that the population dynamics of the dinoflagellate community fluctuate with time. Dinoflagellates include many species inhabiting various aquatic ecosystems. Their nutritional mode (autotrophic, mixotrophic, heterotrophic, symbiotic, or parasitic) and chemical defences may affect their ecological niches (bottom of water bodies: the benthos or the water column: the plankton). Heterotrophic species constituted 49% of the total species. They are also reported to be slightly dominant in marine waters (Gómez 2012). This prevalence of heterotrophic dinoflagellates may be explained by their competitive features. For example, they can be raptors feeding on prey using "temporal gradient sensing" chemotaxis, in which cells move along a chemical gradient in a directed manner toward the higher concentration. As the extracellular concentration of the attractant increases, a corresponding increase in the ratio of net to gross displacement results in an overall movement toward the stimulus (Cancellieri 2001). Moreover, heterotrophic dinoflagellates are major Fig. 4 Scanning electron micrographs showing the diversity and structural complexity of the plastisphere that develops on LDPE collected from different stations in Monastir Bay ◂ grazers of diatoms in the sea and constitute also an important food resource for mesozooplankton (Sherr and Sherr 2002).
All these observations make the explanation of the presence and/or absence of one or more species of dinoflagellate on the plastisphere a difficult mission. Certainly, their flexibility in the trophic mode, their mobility, as well as their life cycle make them an "intelligent colonizer" and a "crucial piece of the puzzle" to understand the interaction between the different organisms sheltering the plastisphere. Therefore, more research on the dinoflagellate community growing on the plastisphere is needed.
Our results indicate that diatoms were present in all plastispheres, with RAs ranging from 0.19% at S8 to 40% at S5. The diatom group showed a high diversity of species observed on the two matrices: 25 species on the plastisphere and 23 species in the seawater (Fig. 5d). Some species were only observed on the plastisphere (Cocconeis placentula, Grammatophora marina, Skeletonema costatum, Striatella unipunctata, Synedra ulna, and Pseudo-nitzschia spp.), while others were only in seawater (Asterionellopsis glacialis, Bacteriastrum delicatulum, Coscinodiscus sp., and Guinardia spp.).
Both Navicula sp. and Nitzschia longissima were observed in both matrices but with different relative abundances. In seawater, the highest RA of Navicula sp. (100%) was observed for station S2. For the plastisphere, the highest RA of Navicula sp. (58.06%) was observed for station S6. Other studies showed the importance of diatoms as dominant colonizers of plastic debris that are exposed to sunlight (Oberbeckmann et al. 2014;Masó Fig. 5 Bar blots of relative abundance (%) of different microalgae groups (to total microalgae) (a), cyanobacteria species (relative to total cyanobacteria) (b), dinoflagellates species (relative to total dino-flagellates) (c) and diatoms species (relative to total diatoms) (d) in plastisphere (P) and sea water (SW) at the eight sampling stations Achnanthes brevipes 6.00 × 10 2 3.00 × 10 2 2.00 × 10 5 1.00 × 10 5 Amphiprora paludosa 6.00 × 10 2 3.00 × 10 2 9.60 × 10 3 4.80 × 10 3 et al. 2016; Michels et al. 2018;Kettner et al. 2019;Amaral-Zettler et al. 2020). It was reported that after two weeks of plastic colonization, diatoms multiply on plastic polymers and can reach concentrations up to 10 times higher compared to the surrounding environment (Eich et al. 2015). The crucial role of diatoms during biofilm formation has already been demonstrated (Patil and Anil 2005). However, the general process of biofilm formation, which starts with electrostatic attraction and repulsion between surfaces and negatively charged bacterial cell walls (Renner and Weibel 2011), may apply to diatoms even though the biological process of diatom adhesion to plastic debris has not yet been thoroughly investigated. In fact, negatively charged diatom cell surfaces have been seen (Konno 1993;Gélabert et al. 2006). The diatoms were more readily observed on the film than the small microplastics. This suggests that the rigidity morphology of diatoms requires a flatter surface area for colonization compared to bacteria (Cheng et al. 2021).
NO NO 2.59 × 10 5 1.30 × 10 5 to total potential harmful cyanobacteria (RA Cya ), the relative abundance of the potential harmful dinoflagellate species in relation to total potential harmful dinoflagellates (RA Din ) and the relative abundance of potential harmful diatoms species in relation to total potential harmful diatoms (RA Dia ). The analysis of the results of the mean relative abundance of total PHM (RA PHM ) in both matrices (Fig. 6) showed that there is a significant difference (p ≤ 0.05) between the stations. On the other hand, PHM are more abundant in the plastisphere than in seawater at all stations. Four of the eight study stations contained plastic debris with plastispheres containing more than 40% PHA. The RA PHM ranged from 1.42 ± 0.26 for S1 to 55.82 ± 3.02 for S8. For SW, the highest RA PHM did not exceed 22.57 ± 1.10% (station S4).
The mean relative abundances of PHM for the plastisphere are remarkably high for stations S8, S7, and S2. They are 97, 26, and 28 times higher than those for seawater, respectively. These sites are characterized by shallow water, low hydrodynamics, and clandestine dumping. Station S1, unlike all others, has low RA PHM for both matrices (1.19 ± 0.36% and 1.42 ± 0.26%, in seawater and plastisphere, respectively). Station S1 is located in the open sea of Monastir Bay, in front of the Kuriat Islands. This station is the deepest (20-25 m depth) and the marine dynamics are quite active, which reduces the amount of plastic floating on its surface and the duration of its stay. Compared to the coastal area, relatively little plastic waste was found during the sampling event in this station. Even the samples that were taken did not reveal a dense plastisphere, which suggests that the debris were just recently thrown out.
The RA G analysis (Fig. 7a) reveals that for all coastal area stations (except S1: offshore station, Kuriat Islands), the most PHM colonizing the plastisphere are potentially harmful cyanobacteria. For station S1, the PHM community profile was characterized by the dominance of potentially harmful diatoms with a RA G of 81.5%. Filamentous and colonial cyanobacteria are frequently the causes of HAB in inland waters. Their interaction with plastic debris in wastewater effluent may influence HAB formation and persistence (Lee 2008). In fact, filamentous cyanobacteria that grow on the plastisphere of floating plastic can aggregate and contribute to the formation of surface scum. Many HAB-forming cyanobacteria boost buoyancy via gaseous vacuoles (Oliver 1994). This reduces the density of the plastic and slows its sequestration in the sediment. In contrast, when attached to neutrally buoyant plastic debris, specialized cyanobacterial cells such as heterocysts and akinetes (which lack gas vacuoles) can increase the sequestration of its polymers in the sediment. Cyanobacterial cells "hitching" on rapidly sinking plastic may provide a new pathway to escape unfavourable conditions and thus may increase the size of the cyanobacterial "seed bank" in the sediment and possibly the frequency and/or severity of future HABs (Yokota et al. 2017).
The potentially harmful cyanobacteria observed in seawater are mostly composed of the species Lyngbya majuscule, except at station S2 where Phormidium sp. dominates with a RAcya of 100%. Whereas, the main potentially harmful cyanobacterium identified on the plastisphere was Phormidium sp. (Fig. 7b). The RAcya of Phormidium sp. was equal to 100% for S1, S5, and S6. The genus Phormidium has been found on the plastisphere of various types of plastic debris from various geographical origins (Zettler et al. 2013;Oberbeckmann et al. 2014;Bryant et al. 2016;Oberbeckmann et al. 2016, Du et al. 2022Dey et al. 2022). Phormidium and related hydrocarbon-degrading taxa molecular signatures were discovered in plastics by two studies: Zettler et al. (2013) on microplastics collected from the North Atlantic and Oberbeckmann et al. (2014) on PET bottles incubated in the North Sea. This raises the intriguing possibility that the cyanobacteria Phormidium that live in the plastisphere actively hydrolyse the plastic rather than relying on inorganic nutrients found in the surrounding environment or released by heterotrophic bacteria (the primo-colonizers of plastisphere) (Harrison et al. 2014;McCormick et al. 2014;Yokota et al. 2017). Phormidium has the capacity to grow large benthic mats in lakes and rivers. Since some strains produce the potent neurotoxic metabolites (anatoxina, dihydroanatoxin-a, homoanatoxin-a, and dihydrohomoanatoxin-a) also known as anatoxins (Gugger et al. 2005;Heath and Wood 2010;Wood et al. 2012). Worldwide, Phormidium mats have been linked to several dog poisonings and deaths (Faassen, et al. 2012). To the best of our knowledge, few studies have demonstrated the toxicity of  Bar blots of relative abundance (%) of different groups of potential harmful microalgae (to total potential harmful microalgae) (a), potential harmful cyanobacteria species (relative to total potential harmful cyanobacteria) (b), potential harmful dinoflagel-lates species (relative to total potential harmful dinoflagellates) (c), and potential harmful diatoms species (relative to total potential harmful diatoms) (d) in plastisphere (P) and sea water (SW) at the eight sampling stations marine Phormidium strains. Williamson et al. (2002) have highlighted the ability of the marine cyanobacterium Phormidium sp. to synthesize a toxin: phormidolide. The latter is very highly toxic to brine shrimp with an LC 50 = 1.5 μM. Phormidolide is one of only a few macrolide-type natural products to be reported from marine cyanobacteria.
The cyanobacterium Lyngbya majuscula has also been observed but with low RA Cya . The highest RA Cya (42.85%) was detected on station S4. The toxicity of marine cyanobacteria Lyngbya majuscula has been observed on various marine organisms: Brine shrimp, Biomphalaria glabrata mollusc, and Chelonia mydes green turtle (Gerwick et al. 1994;Arthur et al. 2008).
In contrast to the seawater, potentially harmful dinoflagellates were found in low abundance (3.21%) on the plastisphere. Prorocentrum lima is the most dominant. It has been identified in three stations (S3, S6, and S7). Three other species were also identified: Alexandrium minutum, Coolia monotis, and Prorocentrum convacum. The LD-PE-associated dinoflagellates we identified could be associated with HABs.
Species of the genus Alexandrium cause paralytic shellfish poisoning in humans. The first evidence of plastic HAB association was described by Masó et al. (2003), who found temporary cysts of Alexandrium taylori adhering to plastic bottles. Other potentially toxic dinoflagellates in our plastisphere samples included Coolia monitis, considered a potential cause of seafood poisoning (Faust 1995;Lenoir et al. 2004), and the genus Prorocentrum, which contains several toxic species, some of which can generate toxins that play a role in ciguatera syndrome.
The second group of potentially harmful microalgae found on the plastisphere is diatoms. Five stations recorded their existence (S1, S2, S6, S7, and S8). The station S1 (Kuriat island) recorded the highest RA G (81.5%), while station S8 recorded the lowest (0.62%). Our findings also showed that there were no potentially harmful diatoms in the seawater matrix. Pseudo-nitzschia spp. were the only potentially harmful diatom observed on the plastisphere (Fig. 7d). Our results also demonstrated the total absence of potentially toxic diatoms in the seawater matrix. Our results are consistent with other work that has shown the dominance of Pseudo-nitzschia spp. in the marine plastisphere (Casabianca et al. 2018;Dudek et al. 2020;Cheng et al. 2021). The diatom genus Pseudo-nitzschia is found all over the world and some of its species generate domoic acid (DA), the neurotoxin that leads to amnesic shellfish poisoning. At least three people were killed by this poison in 1987, which led to a plethora of studies on the physiology, distribution, ecology, and identification of Pseudo-nitzschia species. Since prior assessments in 1998, more understanding has been obtained regarding the fate of DA, including how marine species accumulate it and how light and microbes break it down. A review of the biotic and abiotic elements that affect DA production is done with an emphasis on how recent findings have altered our initial theories regarding the control mechanisms. Recent research has confirmed that silicate and phosphate deficiency cause the formation of DA. However, recently identified triggers for stress, such as high or low copper or iron concentrations, point to DA's potential function in trace-metal chelation (Lelong et al. 2012). Table 4 shows the trace elements in surface sediments with a large difference between the values for contaminant accumulation in the sediment. Sediment has been considered the central accumulator of trace elements in the marine environment (Tessier et al. 1979). The degree of trace metal enrichment in the sediment of the Monastir Bay coastal area can be evaluated using NOAA sediment quality guidelines (SQG) known as effects rangelow (ERL) and effects range-median (ERM) (Long and Morgan 1990;Long et al. 1995). This is the enrichment of trace elements above which toxic effects in benthic organisms can be expected with a frequency of at least 50% (ERM), and below which effects can be expected only rarely (ERL). A toxic effect can be detected for Cd in the central part of Monastir Bay (S6), which is known as the outfall of the agricultural zone where pesticides are used, and in the coastal zone, there are sites for ships and ship repairs with the installation of offshore aquaculture cages. This cadmium accumulation in sediment has been observed in recent decades (Khiari et al. 2022).

Metal accumulation in sediment
The Geoaccumulation Index (Igeo) ( Table 5), calculated using as background values the trace elements estimations of Turekian and Wedepohl (1961), confirms the extreme Cd pollution at this site and strong to extreme pollution in northern coastal areas S3, S4, and S5 (Fig. 8a). For Pb, Igeo shows no or moderate pollution in the central part of the bay.
The middle part is known to be derived from waste from boat painting (S3), industrial waste (S4), and water treatment plants (S5); areas indicated in Fig. 1. The contamination factor is used to evaluate the potential risk of heavy metal accumulation in biological systems (Hakanson 1980). This index is commonly used to evaluate the accumulation of heavy metals (Rutkowski et al. 2020;Yan et al. 2016). This study used this index to assess heavy metals in the plastisphere. The origin of the plastisphere is one of the most important parameters of our study. It should be noted that the LDPE samples were collected from the tidal fluctuation zone, with some of them in contact with the sediment. Because there is no metal enrichment in the sediment, we used it as the reference background value. Only for Cd, this enrichment is considered.
The enrichment of metals in the plastisphere represents a significant accumulation of trace elements with a contamination factor index higher than 6 (Table 4), indicating that most industrial, port, and aquaculture activities at stations S1, S2, S7, and S8 have a high contamination factor. At station S1 there are aquaculture activities; the other stations are affected by urban and port waste. The average contamination factor at all sites has a higher average value for the following sequence of metals: Pb > Cu > Cd > Ni > Zn. Cobalt is not included in this classification because it occurs only in the central part of the bay (S4 and S5).
A comparative view of metal accumulation in the sediment and plastisphere is shown in Fig. 8. Metal accumulation appears to be influenced by waste from urban industrial areas and ports (S3, S4, and S5). A particular process was observed at station S6, where antifouling paint is used for aquaculture cages in this coastal area. According to Singh and Turner (2009), trace elements Cd, Cr, Ni, and Pb have maximum concentrations in the antifouling residues in the coastal sediment, indicating a heterogeneous source of metallic contaminants in the marine environment where cage cleaning and maintenance activities occur.
Metal accumulation in the plastisphere is most common near Ras Dimas and Ksibet El Madiouni, where there is a specific morphological feature with a very shallow water depth, not exceeding 0.5 m depth and with very low hydrodynamics, as explained by Khiari et al. (2022). As shown in Fig. 8, trace elements in the plastisphere have accumulated mainly in this zone, which is characterized by low bathymetry. Due to the marine barrier (Dahr) and Ras Dimas spit, which have canceled the hydrodynamics in this area, these pollutants, brought by the runoff and released by the port and fishing activities, have been deposited on the plastisphere without any mechanical dynamics (Khiari et al. 2022).
In this context, the interaction of metals with another matrix is the main cause of plastic-metal ion interaction in terms of competitive processes. The contact between plastic, other solid phases, and the water phase was subject to both continental and marine influences in the coastal zone. Water chemistry, especially salinity, pH, temperature, redox Table 5 Geo-accumulation index in sediment (Igeo) and accumulation factor (C i f ) in the plastisphere I-geo (sediment) conditions, and other dissolved ions, are the main parameters affecting the properties of the plastic matrix and associated elements (organic or inorganic). An important factor in this complex is the degradation rate and biofouling of polymers, which affect the bioavailability of metals, wettability, and reactive surface area (Binda et al. 2021). Environmental processes have a great influence on the way plastics and metals interact and associate.
The accumulation of plastic debris and trace elements in aquatic and terrestrial environments is a major concern (Bradney et al. 2019). The adsorption of trace elements onto the plastisphere in the marine environment is generally a result of adsorption onto plastic surfaces found in freshwater and terrestrial environments (Du et al. 2022). On the other hand, there is adsorption of plastic debris present in seawater, which is less common as the continental environment is the main source of plastic waste (Guo and Wang 2019). Plastic debris act as vectors to transport trace elements and can increase the chemical exposure of ingesting organisms and the resulting toxicity (Bank 2022).
Interactions between metals and plastic debris occur through a variety of modes, primarily direct adsorption of cations (physical); other forms of association of metals to the plastisphere by complexation on charged sites or neutral regions of the plastic surface; and complexation on charged sites or neutral regions of the particulate plastic surface that will be the site of chemical co-precipitation (Binda et al. 2021). Other types of the metal-plastisphere association include adsorption onto hydrated oxides (Ashton et al. 2010).

Conclusion
This study, which is the first one carried out on this subject in Africa, has shown that the plastisphere can be considered an "extreme ecosystem" and a "new field of study for bioremediation." In fact, it has demonstrated great ability of the epiplastic microalgal community to bioaccumulate trace metals at very high concentrations, despite their toxicity. It is thus a "trace metal sponge" community. Moreover, the plastic substrate, which is a synthetic support of hydrocarbons, can develop a diversified community of epiplastic microalgae with the presence of potentially toxic species. This will result in double toxicity of the plastisphere towards the whole food chain. Cyanobacteria and diatoms are the dominant species in the microalgal community that have been observed on the plastisphere. These typically phototrophic microalgae are well known for playing significant roles in biofilm formation because of their capacity to secrete and produce extracellular polymeric substances (EPS). Isolation and characterization of epiplastic microalgal communities is a prospective research area to investigate the bioremediation potential of these microorganisms because it is well-known that heavy metals and these polymers interact. In addition, it is imperative to start prioritizing and identifying processes that occur in the plastisphere, such as biodegradation or pathogenicity. To accurately interpret what is occurring in nature, future studies must include going "back to the bench" and discovering the mechanisms involved in degradation processes.
In the end, "nothing is lost, nothing is created, everything is transformed" (Antoine de Lavoisier, inspired by Anaxagore de Clazomènes).