A comprehensive sampling effort was conducted in the Kara Sea, resulting in a total of 42 stations strategically distributed across various regions ( Figure 1 ). The sampling included 5 net thrawlings specifically performed in the Kara Gate. Six stations were made in the middle of the western part of the Kara Sea shelf to capture the representative conditions in this area. Recognizing the importance of the Ob-Yenisey river plume, 13 stations were dedicated to this region to assess the impact of freshwater inputs. To explore the northern reaches of the Kara Sea, 3 stations were located beyond the 82N latitude. The northernmost point was at 83N. Six stations were positioned over the St. Anna Trough, a dynamic and hydrologically important area. Finally, 8 stations were placed to the north of Novaya Zemlya in the zone influenced by the Western Novaya Zemlya Current, providing valuable insights into this region.

The collection of microplastics floating on the sea surface was carried out by neuston net (40 x 60 cm opening, nylon material, 0.333 mm mesh size) trawling over the sea surface at low vessel speed (2-3 knots) for about 20 min. 20 cm of the net opening was submerged. The method was recommended for surface water microplastic sampling by AMAP (AMAP, 2021). After trawling, the net was washed from the outside with sea water and the content of cod end was washed off with filtered water (0.45 µm pore size filter) into a clean glass jar with a lid. The jars with suspended matter were transferred to a clean room for further processing. Flowmeter was used to calculate the exact filtered water volume, which varied between 76-616 m³.

Additionally, hydrophysical parameters of the surface water were continuously measured in the flow system at 3 m depth during the cruise. The system was equipped by a thermosalinograph (SBE 21 SeaCAT) that was continuously recording salinity and temperature of the subsurface seawater.

No treatment was carried out prior to the analysis. Collected suspended matter was washed off with filtered water from the jars into a plastic basin for primary visual analysis (Lusher et al., 2020) for the presence of microplastics (>500µm) in the sample. After visual sorting, the sample was washed onto a metal sieve with a mesh of 0.5 mm for final verification, so the size range of microplastics in this study was 0.5 – 5 mm. During the analysis, a lamp with an illuminated magnifying glass was used. Potential plastic particles were stored in Eppendorf tubes before analysis where all particles were photographed on graph paper then weighted for each station (AND GH-252 dual-range analytical balance) and all the found particles were identified by Fourier Transform Infrared Spectroscopy (Perkin Elmer FT-IR ATR Spectrum Two). Measurements were obtained at 4 cm^-1 spectral resolution for the range 4000 to 600 cm^-1. Library matching was performed in the Spectrum 10 software (v. 10.6.2). Each spectrum was compared to several different libraries available at NIVA: PerkinElmer ATR Polymers library, BASEMAN library (Primpke et al., 2018), and several in-house libraries including reference polymers, different textile materials, and potential sources of laboratory contamination. All spectra were manually inspected to ensure that spectra showed characteristic peaks of polymers and the library matches were acceptable (>0.8). Textile fibers were not included in the dataset as it was impossible to control contamination by fibers during sampling and used analysis by visual sorting is not suitable for quantitative extraction of microfibers. Technical (solid) fibers like fishing lines were included into the dataset.

To control and minimise external contamination, the following steps were performed: Only distilled and filtered water was used to wash the equipment and working surfaces. The neuston net was rinsed from outside with seawater before each sampling. The composition of all plastic materials used during sampling and analysis was identified and taken into account during analysis (material of the net, boat, laboratory equipment). For control of airborne contamination, a wetted GF/A filter was exposed in an open glass Petri dish close to the sample during processing and analysed for foreign plastic particles after sample processing. Given that no contamination was observed within the target size range we did not perform any corrections of our data, apart from removing particles which were not shown to be plastic after FTIR. Furthermore, textile fibers were excluded from the analysis due to the challenges of controlling for this contamination during field sampling.

To control the quality of the analysis a recovery test was performed in the laboratory. Standard material of three polymers (polyethylene, polypropylene and polystyrene) were added to the cod ends containing some marine suspended matter. Specifically, 21 particles (0.5 – 5 mm) were added to the samples and processed in the same way as the samples. This was performed in 3 replicates to ensure method validity. All analysis steps were followed according to the field samples, including visual sorting and FTIR measurements. The test showed 100% recovery rate and correct polymer identification.

The Python 3 packages for data analysis (scipy, pandas, scikit-learn) were used to process the raw field data. Several regions in Kara Sea were identified based on hydrophysical properties of water (temperature and salinity) using Ward’s hierarchical clustering method. Main statistics (mean and standard deviation) were calculated for microplastics data within the determined sea regions. The t-test for independent samples was used to quantify the difference between the arithmetic means of the microplastics abundance, weight concentrations and particles sizes.

Wind data were obtained from Global Ocean Hourly Sea Surface Wind and Stress from Scatterometer and Model (European Union-Copernicus Marine Service, 2022). Wind direction and speed were also averaged with hourly frequency, which allowed us to plot wind roses for every sampling site. Satellite MODIS AMSR-2 sea ice concentration was used to map the ice margin during the cruise.

Maps were plotted with QGIS 3.28. Microplastics concentrations were calculated in different units, in items/km², items/m³ and µg/m³, so that they can be easily compared with the results of other studies. Abundance (items/km²) was calculated as follows:

Using the hierarchical Ward method based on the dendrogram, three classes of water were identified ( Figure 2A ):

The temperature and salinity at microplastic stations were plotted on the TS diagram of the cruise ( Figure 2B ). Stations 4-14 were assigned to the class of highly saline warm waters, stations 16-28 to the class of riverine waters, and stations 29-45 to the class of highly saline cold waters. Station 15 was excluded from the classification because its temperature (9.2°C) and salinity (24.4PSU) did not fit into any of the described classes. In further analysis the Kara Gate stations (4-8) were isolated into a separate class. It was done due to the significantly higher number of plastic particles found there (1374 in total for 4-8 stations), differing by 2 orders of magnitude from the other stations. Also, stations from the class of highly saline cold waters were divided into two regions according to the water circulation patterns: stations 29-37 in the Northern Kara Sea including St. Anna trough and further to the Arctic Ocean basin, and stations 38-45, influenced mainly by Western Novaya Zemlya Current and belonging to the Barents Sea ( Table 1 ).

A total of 1679 found particles were identified as plastic, where 1507 items (90%) were<5 cm in size and belonged to microplastics. Fibers (lines) were detected only in the Kara Gate, representing on average 10% of the total microplastic fraction in this area, all other found items were fragments. The abundance of microplastics varied from 0 items/m³ to the maximum value in the Kara Gate 1.53 items/m³. The average value for the entire study area is 0.124 ± 0.383 items/m³. All types of plastics with positive buoyancy have been found, the most common being polyethylene (PE), polypropylene (PP) and expanded polystyrene (EPS). The results are summarized in the Table 1 (in details is given in Supplementary Table S1 ).

The difference between the divided areas of the Kara Sea was found in terms of the number of stations free from MPs. Overall distribution was quite patchy, especially in the river plume and at W. N. Zemlya Current were 23 and 50% of stations were free of MPs, respectively. The most polluted was the Kara Gate and Western Kara Sea – MPs were observed at all stations here. Moreover, most of the particles >5 mm, meso- and macroplastics, were also found in the Kara Gate and Western Kara Sea ( Figure 3 ; Table 1 ).

The results revealed that the Kara Gate exhibited the highest abundance of microplastics (0.93 ± 0.73 items/m³), marking it a critical point for microplastics transport into the Kara Sea. This influx of MPs potentially represents a composite of MPs contained within the Norwegian Coastal Current, branch of North Atlantic Current, affected by inputs from coastal sources along its way. Additionally, contributions may originate from the White Sea and local sources within the coastal zone of the Barents Sea. The Western Novaya Zemlya Current may serve as a significant conveyor of MPs directly from the North Atlantic Ocean. Local sources within the open Barents Sea, that is maritime activity, may also play a crucial role.

The water coming from the Barents Sea spreads across the Kara Sea shelf; some part of it is captured by the Yamal Current that directs it towards extensive Ob-Yenisey river plume. Plume front limits the further propagation of MPs from the Barents sea to the east and contributes to its retention in the Western Kara Sea. Higher abundance of microplastics within Western Kara compared to ambient concentrations were observed before (Pakhomova et al., 2023). While the Ob and Yenisey rivers have been identified as potential pathways for microplastic pollution in the Kara Sea (Yakushev et al., 2021), our results suggest their contributions might not be the predominant sources due to significantly lower (p< 0.05) concentrations (0.008 ± 0.009 items/m³) compared to Western Kara Sea (0.023 ± 0.011 items/m³) and Kara Gate (0.93 ± 0.73 items/m³). Instead, the presence of microplastics may be attributable to intensive maritime activities, particularly in the Gulf of Ob, including shipping, fishing, and oil and gas operations. Only the upstream regions of Ob and Yenisey rivers are populated, which can be a source of MPs pollution. It can be assumed that when transported downstream, plastic particles are fragmented, beached, consumed by biota and sink to the bottom. The main plastic pollution remains in upstream regions of the rivers (Frank et al., 2021).

It is well-documented that the dynamics of the river plume in the Kara Sea, and hence the dispersion of riverine microplastics, besides buoyancy, are largely driven by wind patterns (Osadchiev et al., 2017). Strong southern (offshore) winds in late summer and autumn induces eastward Ekman transport of the freshwater plume and promotes its spreading towards Vilkitsky Strait (Osadchiev et al., 2020). The north-western wind is dominant during the summer period with the “western” type of the river plume propagation. Such wind induces the Ekman transport directed to the west or northwest (Kubryakov et al., 2016). A month before the sampling campaign northern wind prevailed in the Ob-Yenisey river plume area that impeded the spreading of riverine microplastics seaward. These patterns significantly influence the transport and distribution of riverine microplastics, steering them towards certain areas. As such, the prevailing wind conditions play a fundamental role in determining the spread and fate of riverine microplastics within the Kara Sea.

Importantly, we also observed significant differences in the polymer compositions of microplastics from river sources and those in saline waters. The most pronounced case is the predominance of EPS particles (52%) in the River Plume in contrast to other regions dominated by PE. EPS is frequently used for packaging, insulation, and as a cushioning material in the transport of goods. Activities along the river coast, such as construction and waste disposal, might contribute to the direct input of EPS particles into the river system. Due to its low density and porous structure, EPS particles can float on the water surface, making them more likely to be transported downstream the river. The predominance of EPS suggests that land-based activity is a main source of microplastics in the river plumes of the Kara Sea. The divergence in chemical composition of microplastics from river sources correspond to the findings reported by (Yakushev et al., 2021), highlighting the diverse origin and possibly differing environmental impacts of microplastics based on their source.

Northern Kara Sea is affected by waters coming from the rivers, from the North Atlantic via W. N. Zemlya Current, by coastal Barents Sea water entering through the Kara Gate and spreading in western Kara Sea as well as by Arctic Ocean water from the North. Microplastic composition found in the Northern Kara Sea suggests several contributing sources. These include inputs from the Western Novaya Zemlya Current and the Kara Gate, indicated by similar fractions of PP and PE – 77% PE and 18% PP in the Northern Kara Sea and 64, 77% PE and 18, 20% PP in W. N. Zemlya Current and the Kara Gate respectively. Given the large proportions of PP and PE in both the Warm Saline waters and the Western Novaya Zemlya Current, identifying the prevailing source becomes a challenging task. From the other site, presence of EPS (5%) in the Northern Kara Sea can point on influence of riverine waters here as in the river plume EPS was the main polymer type found (52%) while it was negligible in the Kara Gate and the W.N.Zemlya Current. Additionally, drifting ice can capture and transport microplastics over long distances (Obbard et al., 2014; Kanhai et al., 2020; Kim et al., 2021; Mountford & Morales Maqueda, 2021), so this can explain the mere fact of presence of microplastics in the northern Kara Sea beyond 82N.

During winter and extensive sea ice coverage, the movement and dispersion of microplastics are impeded, as the ice acts as a physical barrier that restricts their transport and accumulation.

In recent studies, a substantial presence of microplastics has been reported within the Arctic’s sea ice, reflecting alarming levels of pollution (Peeken et al., 2018; Huntington et al., 2020). This detection of high concentrations raises concerns about the potential release of particles as the sea ice retreats during the summer. Though, these processes were not directly accessed in the present study, the results did not indicate an increase in microplastic concentrations near the ice margins in the northern part of the Kara Sea (stations 29-31).

The highest microplastics abundance (up to 1.53 items/m³) was found in the Kara Gate, suggesting that it is an important source of the microplastics pollution of the Kara Sea.

The hydrological conditions in the Kara Gate are influenced by the dynamics of the Pechora river plume and wind-driven currents. Recent studies revealed regular advection of the Pechora plume through the Kara Gate into the Kara Sea (Rogozhin et al., 2023). Pechora plume can act as a barrier from the south for microplastics coming from the Barents Sea and direct particles towards the Kara Gate or prevent the influx of MPs coming from Barents Sea.

During summer surface water from the Barents Sea carried by Pechora current in the south flows eastward while the water from the Kara Sea (Litke current) flows towards the west ( Figure 1 ). In the summer the Pechora River discharge represents a large portion of the easterly flow, mixed with saline waters of Atlantic origin (Pfirman et al., 1997). North-east and east wind promotes the outflow from the Kara Sea and reduces the inflow of Barents Sea water (Pavlov et al., 1994). In August, when the observations were conducted in this study, the incoming surface current typically increases in width (Pfirman et al., 1997). The stations 4-8 were made in Pechora current according to the decreased salinity of surface water (32.2 – 32.5 PSU) affected by Pechora plume and higher temperature (11.1 – 12.1°C).

Given the relatively higher concentrations of macro- and microplastics in the Barents Sea (Grøsvik et al., 2018; Novikov et al., 2021; Pakhomova et al., 2022). Pechora current can potentially introduce substantial amounts of microplastics into the Kara Sea. Under favourable conditions, the intersection of two opposing currents can generate eddies, creating a local trapping zone for microplastics (van Sebille et al., 2020). This could lead to localized areas of higher microplastic concentration (Pfirman et al., 1997). It is worth noting that, as the Kara Gate is a coastal zone, generally, a higher concentration of microplastics might be expected in this region compared to the open ocean.

Another point is that the Kara Gate, as a part of the Northern Sea Route, experiences relatively significant maritime activity. This high traffic maritime zone is a potential hotspot for the introduction of microplastics into the marine environment due to operational activities, including waste discharge, cargo handling, shipping accidents, and hull maintenance. Ships can discharge microplastics through the dumping of waste and litter, the spilling of cargo that contains plastic materials, or the release of gray and black water that contains synthetic fibers from laundry.

Approximately 80% of the plastic coming from ships or maritime activities originates from fishing ships, especially their gear (Deville et al., 2023). Modern fishing nets and ropes are predominantly made from three types of polymers: polypropylene, polyethylene, and nylon (Lipinski et al., 2020). Within the whole Kara Sea microfibers were only identified in the Kara Gate, and their composition predominantly of PP (polypropylene) strongly suggests an origin linked to fishing and shipping activities.

Currently, data on floating microplastics in the Eurasian part of the Arctic, excluding the findings from this study, are limited to 3 publications (see Table 2 ). The table below also includes data by (Ikenoue et al., 2023) as Abundances (items/km²) were presented there.

Cózar et al., 2017 sampled both micro- and macroplastics, while the Tošić et al., 2020 focused on smaller particle sizes and also included fibers in their data analysis. These methodological differences can lead to higher reported plastic concentrations and hinder direct data comparisons. Nonetheless, a shared trend across all three studies is evident: the highest concentrations are observed in the Barents Sea, decreasing eastward. The average abundance of microplastics in the Kara Sea was found to be comparable to the data presented by (Ikenoue et al., 2023) for the Chukchi Sea.