Quantitative and qualitative evaluation of plastic particles in surface waters of the Western Black Sea

Microplastic abundances have been studied intensively in the last years in marine and freshwater environments worldwide. Though several articles have been published about the Mediterranean Sea, only few studies about the Black Sea exist. The Black Sea drains into the Mediterranean Sea and may therefore significantly contribute to the Mediterranean marine pollution. So far, only very few articles have been published about micro-, mesoand macroplastic abundances in the Western Black Sea. In order to fill this knowledge gap and to decipher the number of plastics on the water surface, 12 samples were collected from surface waters with a neustonic net (mesh size 200 mm) in the Black Sea close to the Danube Delta and the Romanian shore. Organic matter was digested and plastic particles were isolated by density separation. The results of visual inspection, pyrolysis GC-MS (for microplastics) and ATR-FTIR (for mesoplastics >5 mm) revealed an average concentration of 7 plastic particles/m, dominated by fibers (~76%), followed by foils (~13%) and fragments (~11%). Only very few spherules were detected. The polymers polypropylene (PP) and polyethylene (PE) dominated which is in line with other studies analyzing surface waters from rivers in Western Europe as well as in China. Statistical analyses show that the plastic concentration close to the mouth of the Danube River was significantly higher than at four nearshore regions along the Romanian and Bulgarian coastline. This could be explained by plastic inputs from the Danube River into the western part of the Black Sea. © 2020 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

Being located between South-East Europe and Asia Minor, the Black Sea is an important basin for aspects such as ecology, geology, tourism, economic or international affairs. The Black Sea has a catchment basin of 2 million km 2 draining 21 countries (Lehmann et al., 2015) with a total population of ca. 180 million (Lehmann et al., 2015). It is a narrow semi-enclosed basin characterized by a high deposition of multiple pollution forms (Bakan and Büyükgüng€ or, 2000;Topçu et al., 2013;Tuncer et al., 1998) and is often described as a highly degraded ecosystem (BSC, 2007) due to several geomorphologic and socio-economic factors: intense maritime navigation, high discharge of polluted freshwater from several large rivers (Danube, Dnieper, Bug, Dniester, Don, Kuban, Rioni), numerous industrialized harbors and cities surrounding the sea, intense fishing activities (FAO, 2015), strong tourism, etc. Heavy metal pollution (Duliu et al., 2009;Sarı et al., 2018), organochlorine pesticides (Ozkoc et al., 2007) and water and sediment quality (Akbal et al., 2011) have been studied intensively; however, only few studies about micro-(<5 mm), meso-(5e25 mm) and macroplastic pollution (>25 mm; GESAMP, 2019) in the Western Black Sea have been published. Topçu et al. (2013) studied marine litter, Suaria et al. (2015) the pollution of sandy beaches and surface waters of the Southwestern Black Sea, Moncheva et al. (2016) marine litter in the bottom sediments of the Western Black Sea and Simeonova and Chuturkova (2019) the abundance of marine litter along the Bulgarian Black Sea coast. Popa et al. (2014) investigated microfibers in marine habitats and Aytan et al. (2016) examined microplastic particle concentrations in surface waters along the Southwestern coast of the Black Sea. Furthermore, € Oztekin and Bat (2017) analyzed microplastics in the water column and surface waters of the Southern Black Sea and Aytan et al. (2018) the abundance of microplastic particles in copepods (southwestern part). Microplastics in beach sediments were studied for the first time by Popa et al. (2014) at five sites of the Romanian Black Sea coast.
To add to the current understanding of plastic pollution in the Black Sea, this study has the goal to evaluate plastic concentrations at several locations of the Western Black Sea. Considerable quantities of plastics (4.2 t/day, Lechner et al., 2014) together with sediment and water are transported via the Danube River, especially during the flood season. It is assumed that the north-south surface currents (Shapiro, 2019) strongly contribute to a distribution of these particles not only nearby the southern area of the Danube Delta, but also further south, along the Romanian and the Bulgarian coast. Berov and Klayn (2020) describe similar plastic particle loads as the ones reported in the present study.

Study area
The Western Black Sea from the Danube Delta south to the Bulgarian border was selected for a comprehensive evaluation of surface water in the Black Sea. The study represents an initiative for evaluating the plastic particles in the surface waters of the Western Black Sea, with respect to both studies in the Southern Black Sea area (Aytan et al., 2016(Aytan et al., , 2018 and Guidelines for the monitoring and assessment of plastic litter and microplastics in the ocean (GESAMP, 2019). Possible sources of plastics were taken into account such as the mouth of the Danube Delta (branches Sulina and Sf. Gheorghe) which discharges ca. 6500 m 3 freshwater per second into the Black Sea (Stancik et al., 1988) as well as main harbors and cities (Constanta and Mangalia) (Fig. 1). For investigating the pollution degree, sampling areas with a high ecological importance (e.g. biosphere reserves) were also targeted: the Danube Delta coast (proximal waters) and areas located nearby tourism resorts predominantly situated between the cities of Constanta and Mangalia.

Sampling
Samples were collected during one campaign in August 2018 organized by NIRD GeoEcoMar by using the research vessel RV Mare Nigrum (see supplement S1 for coordinates). Sampling areas were chosen on the one hand as being close to possible sources of plastics and to tourism resorts and on the other hand close to natural areas (Luo et al., 2019;Siegfried et al., 2017) (Fig. 1). Six samples were taken near the Romanian coast 2e5 km from the shore, four samples 10e30 km and two other samples 46 km and 115 km east of the shore (Fig. 1). The floating plastic particles were sampled with an attached neustonic net (HydroBios, mouth opening: 70 Â 40 cm, length of net bag: 260 cm) that was hooked onto a side crane. The net mesh size (200 mm) is comparable to the study by Aytan et al. (2016) who collected samples from the top water layer (25e30 cm). The surface water (upper 20 cm) was sampled for 5 min over a distance of about 300 ± 100 m at a speed of about 1.7 knots. The filtered water was calculated with a manual flowmeter fixed to the opening of the net (see Scherer et al., 2020). The water samples were transferred into pre-cleaned glass jars, using distilled water for rinsing, and then frozen at À20 C. Between each sampling, nets were cleaned with distilled water to prevent any contamination from the vessel.

Sample preparation
The preparation process was performed at the German Federal Institute of Hydrology. The samples were transferred into individual pre-cleaned glass beakers and organic matter was digested using a reagent composed of 10 M potassium hydroxide solution (KOH) and hydrogen peroxide (H 2 O 2 , 30%) (1:1; 30e50 mL per sample; see Ehlers et al., 2019;Scherer et al., 2020). After being agitated for 5e7 days, samples were neutralized with formic acid (3.89 mL per 10 mL KOH). During the agitation, the glass jars were covered with parafilm to prevent airborne microplastic contamination. Microplastic particles were isolated in a separation funnel with potassium formate (CHKO 2 ) solution (3.6 g/mL of the sample) for reaching a final density of 1.6 g/mL. This is an appropriate density as most plastics have a lower density and will therefore float in that solution. After 3e4 days, pressure filtration (Sartorius Stedim, G€ ottingen, Germany) was conducted on glass fiber filters (GE Healthcare Life Sciences, Whatman, GF/D Cat. No. 1823-047, diameter: 47 mm, pore size: 2.7 mm). The filters were stored in individual aluminum bowls and covered for protection.

Identification of micro-and mesoplastic particles
Micro-and mesoplastic particles were identified by means of optical (visual identification), spectroscopical (FTIR) and spectrometrical analysis (pyrolysis GC-MS). ATR-FTIR was chosen for all mesoplastics and pyrolysis GC-MS for selected microplastic particles.

Visual identification
Optical analysis for identifying plastic particles was performed with a digital microscope (Keyence VHX-2000, Osaka, Japan) equipped with 50 and 200 x magnifying lenses. In order to identify particles, color and shape of particles were analyzed (see criteria by Nor en, 2007). For identification of microplastics, all particles with a size of 200 mm to 5 mm were taken into account, with the upper limit after Arthur et al. (2009). Additionally, we quantified particles bigger than 5 mm that were identified in low amounts, framing them in the mesoplastic category (5e25 mm; GESAMP, 2019). Particle abundance was based on filtered water and the amount of identified plastic particles (in particles/m 3 ).

Pyrolysis GC-MS
Isolated single particles or fibers were identified by pyrolysis GC-MS using the system described previously (Dierkes et al., 2019). A single particle was pyrolyzed at 600 C using a split ratio of 1:20 for fibers and 1:50 for particles. For detection, an Agilent MSD 5977B (Santa Clara, CA, USA) in scan mode (40e800 Da) was used. Identification was performed by comparison of the resulting pyrograms with those from the library F-Search 3.4 (Frontier Laboratories, Saikon, Japan; Tsuge et al., 2011) containing >1000 polymers.

Fourier-transform infrared spectroscopy (FTIR)
All mesoplastics with a size >5 mm were measured using the platinum ATR unit (equipped with a diamond) of the Fouriertransform infrared spectrometer (FTIR; Vertex 70, Bruker, Ettlingen, Germany). Analyzing plastic polymer types with methods such as Fourier-transform infrared spectroscopy (FTIR) is essential as up to 70% of particles may be misidentified as plastics during visual analysis (Hidalgo-Ruz et al., 2012). Therefore, in our study representative red, gray, black, white and transparent (colorless) particles were analyzed using an FTIR. The FTIR measurements were performed in attenuated total reflectance mode (ATR) in a wavenumber range of 4000e370 cm À1 with 8 co-added scans and a spectral resolution of 4 cm À1 . The obtained spectra were compared with the Bruker spectral database using the software OPUS 7.5 (Bruker). Only those particles which had a spectrum with a hit quality above 700 were considered as plastics (i.e. Bergmann et al., 2017).

Quality assurance
In order to prevent sample contamination from e.g. dust, the working space was cleaned with ethanol each time before work started. Glass and stainless-steel materials were used for the laboratory work. Moreover, cotton lab coats were worn. During sample preparation, two laboratory blanks were conducted, using same amounts of potassium hydroxide, hydrogen peroxide, potassium formate, formic acid and ultrapure water, which was digested the same way as the other samples and filtered. Correction was applied to each filter. Also, all samples were covered with aluminum foil/ parafilm throughout the chemical digestion and storage.

Statistics
Plastic concentrations were analyzed with the Statistica 10 software. Tentative plastic concentrations (with the exception of the unreplicated locations CT04 and MA02) were examined using a balanced ANOVA that compared the mean plastic concentrations between the Danube Delta and the nearshore locations (two locations paired: Danube Delta: SU01 & SG01, nearshore 1: SG04 & SG05, nearshore 2: PO01 & PO04, nearshore 3: CT01 & TZ18, nearshore 4: MA05 & MA08). For the confirmation of all ANOVA assumptions (i.e. variance homogeneity, normality), Cochran C and Shapiro-Wilk-test were used. Moreover, we ran the Tukey Honestly Significant Differences (HSD) test to check which regions differ in plastic load.

Numerical abundances of plastics in the Western Black Sea
All potential microplastic particles on the filters were visually examined from 200 mm to 5 mm. In total, 3289 particles were counted including 22 particles >5 mm. In all 12 samples, plastic particles were detected (see supplement S2). The results reveal an average concentration of 7 plastic particles per m 3 . Concentrations ranged from 1.3 particles/m 3 (PO04) to 18.6 particles/m 3 (SU01) (Figs. 3 and 6). The most dominant form were fibers followed by foils and fragments. Only few fiber clumps and spherules were found in the samples. Of those, 74.6% were identified as fibers, 1.5% as fiber clumps, 12.7% as foils, 11.1% as fragments and 0.1% as spherules (Fig. 4).
Along the western coast of the Black Sea, plastic concentrations in the surface waters were the highest close to the Danube Delta (Fig. 6, SU01: 18.6, SG01: 16.1 particles/m 3 ). The two samples further southeast are characterized by an evenly low amount (SG04: 5.1 particles/m 3 , SG05: 5 particles/m 3 ) and were similar when compared to the other nearshore regions (PO01 and PO04, CT01 and TZ18, MA05 and MA08). Along the coast to the south close to Gura Portitei, 3.3 particles/m 3 were counted (PO01). Several kilometers to the east, the concentration decreased to 1.3 particles/ m 3 (PO04).

Contamination control
Contamination during laboratory analysis was detected on the two blank samples (one for a set of six samples) with an average of 53 ± 20 fibers (>200 mm) per filter. The particles counted on the blank filters were subtracted from each corresponding sample analyzed, taking into account color and morphology of particles.

Statistical analysis
Five regions were differentiated in the ANOVA. The two samples CT04 and MA02 further to the east of Constanta and 115 km to the east have not been integrated in the statistical analysis as they were not replicated. The ANOVA revealed that there was a significant difference between the five analyzed regions (F(4,5) ¼ 22.05, p ¼ 0.002) and the Tukey HSD test showed that plastic levels close to the Danube Delta were significantly higher than at the four nearshore locations (p < 0.05). However, the nearshore locations did not differ in plastic levels (p > 0.05).

Characterization of polymer types
In total, 93 particles out of 3289 (2.83%) particles were analyzed by means of ATR-FTIR (22 particles) and pyrolysis GC-MS (71 particles). 75% of all particles could be identified as polymers (Fig. 5).
All particles (22) with a size >5 mm were measured with the ATR-FTIR. The analyzed plastics were mainly made of PE (20 particles), but we also found PP (one particle) and one red particle belonging to the polyester family (PES; Fig. 2). 71 particles were measured with the pyrolysis GC-MS. For a representative analysis, the most common forms and different colors were chosen for analysis. The most abundant polymer was polypropylene (PP, 30 particles), followed by polyethylene (PE, 8 particles), polystyrene (PS, 4 particles), polyacrylonitrile (PAN, 4 particles) and polyamide (PA), polyacrylate (PAR) and polyester (PES) (one particle each). 22 particles could not be identified (Fig. 5). Thus, ca. 70% of all particles measured with the pyrolysis GC-MS could be verified as polymers. The fibers made up~46% of all identified particles. 75.8% of the identified fibers were made of PP (25), 12.1% of PAN (4), 9.1% of PS (3), and 3% of PA-6 (1).

Characterization and distribution of plastics in the Western Black Sea
Being a primary hotspot for plastic pollution due to the Danube's freshwater influence, the Western Black Sea and the environs of the Danube Delta were identified to have the highest abundance in plastic particles (16e19 particles/m 3 ), whereas the southern sampling sites revealed lower concentrations (2.7e10.3 particles/ m 3 ). The decrease in plastic particle concentration from the Danube Delta to the south along the Black Sea inner shelf has been confirmed by the balanced ANOVA and the Tukey Honestly Significant Differences (HSD) test. However, the statistical analyses showed that plastic loads in the four nearshore regions did not differ from another.
The northernmost region is under the pressure of a relatively high Danube freshwater input, especially during the flood seasons (spring and autumn) (Stancik et al., 1988). Samples were collected in a period with low precipitations (August). Nevertheless, highest concentrations of plastics were identified nearby the Danube mouths (16.1e18.6 particles/m 3 ; SG01 & SU01), probably as a result of a high plastic transport via the river. Further to the east, the abundance of plastics is presumably decreasing due to the N to S currents (Shapiro, 2019).
The second area, located in front of the Gura Portitei resort, is characterized by moderate concentrations of plastics (1.3-3.3 particles/m 3 ; PO04 & PO01). A difference is observed in plastic morphology: foils and fragments are the most commonly encountered forms in the PO04 sample, while the northern samples are dominated by fibers. This pattern could be explained by differences in current regimes.
The harbor city of Constanta is one of the major ports of the Black Sea with a high frequency of ships and many touristic activities along the coast. For the environs of Constanta, Suaria et al. (2015) reported that almost 90% of all findings floating on the surface were plastic waste. Surprisingly, in this 3rd area (CT01 & TZ18), the amount of plastics detected nearby urbanized areas is lower than in the environs of the Danube influence area (3e6 particles/m 3 ). Although several coastal areas north of Constanta are framed in protected natural areas, there was no difference in plastic load when compared to nearshore regions close to the cities Constanta and Mangalia. This might be explained by the wind regime, sea surface currents and wave movements (Shapiro, 2019) and shows how widespread plastic particles were. Nevertheless, in the study area, the surface currents may cause a north to south movement of the litter, with a transport of particles along the shore that implies an accumulation of Danube sourced plastics on Romanian and Bulgarian shores and in coastal surface waters (Berov and Klayn, 2020).
Closer to the Bulgarian border, the 4th region has been identified with high quantities of plastics (7.3e10.3 particles/m 3 ; MA05 & MA08). The increased plastic pollution close to the Bulgarian border is likely caused by litter input from Mangalia harbor or tourist resorts, indicated by the high abundance of fragments and plastic pieces found at the southernmost MA05 and MA08 sampling locations. It can be assumed that the influence of the Danube River is still visible at the southernmost locations which is indicated by some alike particles with same color/morphological characteristics as in samples from northern locations. The limited number of samples only gives first insights into the plastic pollution of the Western Black Sea. More detailed information about plastics in surface water will be obtained with a larger sampling network. This might help to develop further models of source to sink paths of plastics and, especially, will provide information on their landbased sources. Mesoplastics were surprisingly only found in two samples, close to the coast of Mangalia and in the nearshore region in the proximity of the Danube Delta. The current regime of the Black Sea, generally described as counterclockwise (Oguz et al., 1993), is responsible for a high dissemination rate of plastics and pollutants. The particle abundance can also be associated with surface currents of the Western Black Sea, described as a principally north-south movement, subsidiary eddy-like currents, of both water body and floating particles. For the Romanian coast eddy-like currents were described (Korotaev et al., 2003;St anic a and Chuturkova, 2007) that influence the distribution pattern of plastics in areas close to the coast.
Foils were identified in low amounts, with local exceptions (PO04: 47.6% foils; MA05: 22.7% foils; SG04: 19.3%). These particular anomalies could occur as a result of the eddy currents' presence (plausible in areas within a distance of 0e3 km from the shore, MA05), due to the presence of upwelling phenomena (Mihailov et al., 2012) or anthropogenic factors such as oil platforms present nearby sample location PO04.
Although only few particles have been measured with FTIR and pyrolysis GC-MS, a representative selection of types and colors has been analyzed. The results reveal a high abundance of the lowdensity polymers PP (density 0.89e0.91 g cm À3 ) and PE (density 0.93e0.98 g cm À3 ) (Avio et al., 2017). PP and PE easily float on the water surface and are the most produced polymers worldwide (PlasticsEurope, 2019). This has been confirmed by other studies in Europe, e.g. the River Elbe, Germany (Scherer et al., 2020), Antuã River, Portugal (Rodrigues et al., 2018), a tributary of the Thames River, Great Britain (Horton et al., 2017) and Qin River, China (Zhang et al., 2020). Supplementary visual discrimination of plastics (see supplement S3) reflects the abundance of the observed colors. The predominant particles are represented by black (46.6%), followed by white (24.6%) and blue (16%).

Differences between the Danube Delta and the city of Constanta
The main difference in particle morphology between the northern and eastern study area and the southern area close to the shore and Constanta (especially MA05 and MA08) are represented by the differences in concentration of fibers vs. fragments (see Fig. 6). While areas of high influence of the Danube (like those in front of the Delta and further south, south-east) are characterized by fibers as a major component of plastics, nearshore samples from the southern study area (TZ18, MA05 and MA08) reveal higher percentages of fragments, pieces and foils that might derive from land-based sources (e.g. tourism resorts, harbors). Regarding the polymer characterization of particles, a higher concentration of PP was observed in the northern area (SU, SG, PO, CT) and a higher polymer diversity in the southern region (TZ and MA), composed mainly of PE.

Sources
In the countries surrounding the Black Sea, Jambeck et al. (2015) estimated that up to 1 million tons of mismanaged plastic waste could be entering the Black Sea from all surrounding countries. Sources of plastics are most often land-based originating either from urban areas, rivers, sewage or tourism along the coasts (Wang et al., 2018). Thus, populated areas along the Black Sea shore are possible sources (Volckaert et al., 2012;ARCADIS, 2013). Especially mismanaged waste and dumping along the coast and in the sea were mentioned by several authors (Topçu et al., 2013;UNEP, 2005). From 1996 to 2005 marine litter input decreased in marine and coastal areas (UNEP, 2009). However, in the report it is also mentioned that solid waste from landfills may be released spontaneously into the sea and contributes to a growth in marine litter. UNEP (2009) listed primary sources of marine litter in Romania including municipal garbage/sewage, followed by fishery, marine transport, recreational activities along the coast, ports and industry.
Constanta is one of the most important harbors in Romania (Suaria et al., 2015). Most of the ships directly come to the port via the Mediterranean Sea (Davis, 2018). Thus, the most abundant ship traffic occurs south/southeast of the city and close to Mangalia.
Another important factor for plastic input is the Danube River. Topçu et al. (2013) reported that plastics were the dominant macrodebris in the Black Sea (ca. 47%), potentially originating from riverine inputs. Via the Danube River Delta, several tons of plastic waste are transported into the Black Sea every year. Only some studies are available. Lechner et al. (2014) detected an average of ca. 316 plastics per 1000 m 3 for the Austrian Danube between Vienna and Bratislava (500 mm net used). The majority of items originated from the industry (pellets, flakes and spherules). The authors estimated a plastic input into the Black Sea of ca. 4.2 t per day, thus ca. 1500 t per year. However, van der Wal et al. (2014) only estimates an input of up to 500 t per year, mainly due to the input of the River Siret. Lebreton et al. (2017) took into account 184 dams along the stretch of the Danube. Unfortunately, the origin of plastics in our study was not identifiable. Therefore, a reliable identification of the input sources of plastics was not possible. However, we assume that the PP fibers that dominated our samples could have derived from commonly used PP ropes (Welden and Cowie 2017).
Regarding plastic morphology, most of the particles were identified as fibers (74.6%), confirming the idea that fibers float for a longer period (de Haan et al., 2019). Fibers usually have a lower settling velocity than other plastic forms. This has been shown by Waldschl€ ager and Schüttrumpf (2019) and Khatmullina and Isachenko (2017) for freshwater and by Bagaev et al. (2017) for the marine environment. Fibers may either derive from e.g. clothing or ropes (such as the analyzed PP fibers) and may originate from wastewater or shipping activities (Hidalgo-Ruz et al., 2012).
Considering the most abundant morphology of particles (fibers and foils), we assume that primary plastic objects that could be considered as generating sources of micro-and mesoplastics are black/white plastic bags and textiles. This is supported by the fact that larger particles (mesoplastics) were chemically identified as PE e the primary polymer used in manufacturing plastic bags.
Furthermore, some plastic particles close to the Danube Delta had different shapes than the plastic particles in the southernmost study area. It can be assumed that the missing plastic types either sank to the bottom of the sea due to biofouling (Wang et al., 2016) or were concentrated on sandy shores as a result of the wind regime and nearshore eddy-like currents (St anic a and Chuturkova, 2007) which are present south of the Danube Delta.

Limitations
A limitation of this study is the identification of analyzed particles. Only 2.83% tentative plastics were measured by means of ATR-FTIR and pyrolysis GC-MS with verification as polymers of 75%. An overestimation of tentative plastics is therefore well probable. However, the measured particles were chosen on the basis of representability of forms and colors. In future studies, more measurements have to be conducted.

Comparison with other studies
Few data about plastics in the Black Sea exist until present. Aytan et al. (2016) investigated microplastics in zooplankton samples with a 200 mm net and found predominantly fibers in the samples (between 600 and 1200 microplastics/m 3 ), many more than in our study. However, this is in line with our finding of a large abundance of fibers in the samples. Aytan et al. (2016) counted the particles with a binocular and did not verify them analytically. Thus, an overestimation is well probable. € Oztekin and Bat (2017) investigated the southern Black Sea and found more microplastics in the water column (24 ± 26 microplastics/m 3 ) than on the water surface (~2.7 ± 2.3 microplastics/m 3 ). Only particles >300 mm were included in this study. The findings are lower than the ones in our study (1e19 microplastics/m 3 ) which can be explained by the larger mesh size.
Via the Bosporus, the Black Sea is connected with the Mediterranean Sea where many microplastic studies have been conducted.
In few studies, nets with a mesh size of 200 mm were used. Therefore, the results are difficult to compare. Only studies in the Ligurian Sea, Western Sardinia and Ligurian and Sardinian Sea were conducted with 200 mm nets. Concentrations of 0.31 ± 1.17, 0.17 ± 0.32 and 0.62 ± 2 microplastics/m 3 were found, respectively (Fossi et al., 2012(Fossi et al., , 2016Panti et al., 2015). Other studies (net mesh size: 333 mm) in the Mediterranean also revealed low concentrations of~0.116 particles/m 3 (Collignon et al., 2012).
Other river deltas may also be compared to our study. Fibers were the most abundant form, but microplastic concentrations were low (3.5 ± 1.4 microplastics/m 3 ). Zhao et al. (2014) studied the Yangtze estuary with a 333 mm net and found large differences between the estuary (4137.3 ± 2461.5 particles/m 3 ) and the sea (0.167 ± 0.138 particles/m 3 ). Therefore, it is possible that a large part of microplastics remains in the estuary and is deposited on the river banks or found in sediments (Gonz alez et al., 2016).

Conclusion
In order to fill the gap and to contribute to the knowledge on plastic pollution, this study presents a first account of abundance and composition of plastic particle concentrations in surface waters of the Western Black Sea, in the eastern environs of the Danube Delta and along the Romanian and Bulgarian coasts. The results reveal plastic concentrations (micro-and mesoplastic) of surface waters with 1e19 particles/m 3 and an average concentration of 7 particles/m 3 across 12 study sites. The most dominant forms were fiber and fiber clumps (76.1%), followed by foils (12.7%), fragments (11.1%) and spherules (0.1%). PP and PE were the most abundant polymer types, confirming the hypothesis that low-density polymers mostly float on the water surface.
The results suggest that PE foils (including the mesoplastics) may originate from plastic bags while PP fibers may derive from ropes and fishing nets.
The highest plastic abundances occurred close to the mouth of the Danube River, probably originating from the Lower Danube basin. The concentrations south of the Delta and along the coast were significantly lower and revealed no significant difference between the four analyzed nearshore regions. Thus, it can be assumed that high amounts of the plastic are transported via the Danube River into the Black Sea. It is anticipated to have surmises regarding the source to sink approach and transport pathways that could improve uncertain models of the plastic particles' fate in the aquatic environments. Considering the limited number of sampling sites, we suggest to continue analyzing the studied area taking into account a more complex sampling network. This study provides insights into plastic pollution of the relatively understudied Black Sea and showed that more plastic particles are present in the Western Black Sea, especially in the eastern environs of the Danube Delta, compared to other studies of the Black and the Mediterranean Sea.

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.