Microlitter in Fish and Benthic Invertebrates of the NE Baltic Sea: Abundance, Composition and Bioindicators

2.1 Studied species

To allow for holistic assessment of the microlitter contamination to the coastal ecosystems of the Estonial marine areas of the Baltic Sea, organisms from several taxonomic groups were sampled. To assess the ingestion of microlitter by fish the most common fish, both from an ecosystem perspective as well as from commercial importance, were investigated. In order to elucidate the occurrence of microlitter in the digestive system of different trophic levels of fish demersal predatory feeders, such as perch (Perca fluviatilis L.), pike-perch (Sander lucioperca L.), and smelt (Osmerus eperlanus L.), benthic feeders, such as flounder (Platichthys flesus sensu lato), round goby (Neogobius melanostomus Pallas), bream (Blicca bjoerkna L.), turbot (Scophthalmus maximus L.), eelpout (Zoarces viviparus L.) and omnivorous cod (Gadus morhua L.) and pelagic-feeders, such as Baltic herring (Clupea harengus membras L.) and sprat (Sprattus sprattus L.) were sampled.

Macrozoobenthic organisms evaluated in this study consisted of four commonly found bivalves in Baltic Sea - blue mussel (Mytilus trossulus Gould), lagoon cockle (Cerstoderma glaucum Bruguière), Baltic macoma (Macoma balthica L.) and sand gaper (Mya arenaria L.). Also, omnivorous crustacean Harris mud crab (Rhithropanopeus harrisii Gould) and predatory benthic isopod crustacean Saduria entomon L. were analysed.

2.2 Sample collection

Fish samples were collected from multiple sites in NE Baltic Sea - Gulf of Finland, NE Gulf of Riga (Pärnu Bay), and western Estonia (including eastern Baltic Proper, archipelago area, and western coastal area of the mainland) according to a regionally coordinated methodology [45] (Figure 1). Bivalves were collected from Gulf of Finland and coastal sea of western Estonia between April 2019 - September 2020. Samples of the invasive species R. harrisii, which are currently present only at Pärnu Bay and adjacent area in Estonian waters, were collected during diving between May 2019 - July 2019. Mussels attached to hard substrate were collected during diving by carefully dislodging organisms by hand. Ekman bottom grab was used for sampling infaunal bivalves on soft sediments. Benthic isopod Saduria entomon were collected as a by-catch of gill nets only from Gulf of Finland in November 2019. Fish P. fluviatilis, S. lucioperca, P. flesus, S. maximus, Z. viviparus, G. morhua, N. melanostomus, O. esperlanus, C. h. membras and S. sprattus samples were collected with net series conducted from all sampling sites between April 2019 - August 2020. P. fluviatilis, S. lucioparca, B. bjoerkna and N. melanostomus samples from Pärnu Bay were collected with bottom trawl between April 2019 - June 2020. Sampling duration was kept as short as possible reducing the exposure of the animals to the fishing gear. Hauls were performed using a standard towing speed of 2-3 knots, with a standard haul duration of around 30min. Hauls were conducted at depths ranging from 5 to 9m. Gill nets were chosen for sampling fish in shallow waters and hot spots (harbour, river mouth, etc.) with fishing time of approximately 12 hours.

1.jpg

Figure 1. Distribution of collection sites in Estonian marine area

Note: The extent of the western Estonia is indicated with a circle

The diversity in number of analysed individuals per site depended on the abundance of the particular species in different areas according to their habitat preference. Species/size selection was optimized for regional comparison and, wherever possible, overlapping species were chosen in adjacent areas. Altogether the number of analysed individuals was ~1300 (~500 fish and ~800 benthic invertebrates) and varied between 9 and 168 per fish species and between 68 and 245 individuals per species of macrozoobenthic invertebrates (Table 1).

Table 1. Number of analysed individuals, sampling stations, sampling areas

Note: PB - Pärnu Bay; GF - Gulf of Finland; WE - western Estonia; and total length (cm); and total weight (g) of the specimens; s.d.: standard deviation

Table 2. Species analysed for ingested microlitter in the Estonian coastal sea

Note: Min, max and mean total length (cm) and weight (g) are shown; s.d.: standard deviation

2.3 Sample analyses

All specimens were packed in foil directly on board and the samples were immediately transported to the laboratory and frozen at a temperature -20℃. Defrosting was performed at room temperature for microparticle extraction and analysis. Wet Weight (WW) of each individual was weighed with micro weighing scale (accuracy 0.1mg). Prior to the analyses, individual specimens were rinsed with ultrapure Milli-Q water to avoid secondary contamination of the contents. Before packing, fish were inspected for any disease and those showing signs of feeding or regurgitation were rejected. Thereafter the length and weight of each fish were measured directly on board. The total length (i.e., from the tip of the snout to the tip of the longer lobe of the caudal fin) were measured (up to the nearest mm). The gastrointestinal tract of each fish was individually extracted from the mouth to the cloacae. All fish with completely empty stomach were excluded from the analyses. Length of bivalves were measured with a digital caliper (accuracy 0.01mm). Soft tissue and cartilage of the bivalves were segregated from the shell, tissue was rinsed with Milli-Q ultrapure water before weighting. Minimum, maximum and mean total length (mm) and weight (g) of investigated species are shown in Table 2.

For further analysis, the fish gastrointestinal tracts and bivalve soft tissues were treated by adding potassium hydroxide (KOH 10%) approximately three times the total tissue sample volume to degrade natural organic matter. Samples were incubated in oven (60℃, 15min) to increase digestion speed and placed on the shaker and rotated at orbital rotational mode with a speed of 130rpm for 12-24h. Depending on the amount this treatment was repeated several times as necessary, i.e., until all organic matter was dissolved. Distilled water was added to remaining suspension and stirred. The suspension was filtered on fibre glass membrane (Whatman GF/F, 47mm, 0.45µm porosity) with a vacuum pump under airborne contamination free conditions (laminar flow). Glass funnel above the membrane was rinsed with ultrapure water after filtering each sample. The membrane was placed on a Petri dish, covered by glass top. The number and position of microlitter on the membrane was detected before opening the dish. A stereomicroscope (Olympus SZX10) paired with a camera (Olympus) was used to detect ingested microlitter. To distinguish between plastic and organic material hot needle test was used. Hot needle causes the plastic to melt and deform [46, 47]. The sex of the crustaceans was determined, and the width of the carapace was measured in laboratory. The carapace was opened under the stereomicroscope and all intestines were extracted, placed on a Petri dish, and examined visually under stereomicroscope.

Analyses were based on the suggestions given in Guidance on Monitoring of Marine Litter in European Seas [48] and Monitoring Micro-Litter Ingestion in Marine Fish: a harmonized protocol for MSFD and RSCs areas [49]. All litter items were measured and photographed. To ensure the reliability and validity of the methods, standard procedures were followed during all steps of sample analyses. To reduce observer bias, ingested litter items were regularly double-checked by a second person immediately during analyses or later by using photographs.

2.4 Airborne contamination

To reduce airborne secondary contamination and cross contamination several procedures were applied during sample collection and analyses. This included controlling contamination by working plastic-free and including blank measurements. Blank control was performed regularly at every step of analyses by (a) adding similar volume of KOH 10% as that used in the samples containing the target component in a beaker without samples, (b) placing a damp filter paper in a petri dish in the working area during the visual examination of the samples. The controls were given the same full treatment as the studied specimens to assess any airborne contamination. If the blank was contaminated, then microlitter items with similar characteristics (shape, color, polymer type) were excluded from the results as proposed by Avio et al. [47]. Moreover, to ensure limited levels of contamination during analyses the window/air conditioner was closed, and personnel reduced in the laboratory. Additional contamination originating from researchers clothing was avoided by solely wearing 100% natural fiber clothing, such as cotton. All the equipment was cleaned with ultrapure Milli-Q water before each sample analysis. Lab surfaces area were cleaned with 70% alcohol, also for microscopic inspection area was cleaned before analysing samples in petri dishes. Results of the accuracy tests confirmed that the background contamination was minimal and negligible.

2.5 Data analyses

For the ingested microlitter five indices were used in this study as follows:

(1) Total frequency of occurrence (FO%) by species and area - percentage of individuals with microlitter (microplastics and other microlitter) items detected from the gastrointestinal tract to total number of examined specimens.

(2) Average microlitter abundance - the arithmetic mean of microlitter items found from the intestinal tract of the examined individuals and species to the total number of examined specimens.

(3) Percentage number of items for size class by species. The size classes proposed by Valente et al. [50] are considered the most suitable for monitoring purpose [49] was used in current study: i) 100 - 330µm, ii) 330µm - 1mm, iii) 1 - 5mm. The size is given as the largest distance between two spots for all fragmented particles (also for clumps of fibers), for all long and elongated particles (e.g., fibers) the length of the particles is given.

(4) Percentage number of items for material and shape category by species: fiber (straight, clump, flexibility, plastic or natural); fragment (plastic, pellet, granule, film, foam, paint, metal, glass, wood) [49].

(5) Percentage number of items for color by species [49]. R programming language version 4.0.3 [51] was used for data preparation, analysis and graphics. The differences were considered significant at p<0.05. The data for the abundances of microlitter in the organisms were expressed as the mean ±s.d. Differences in the number of ingested microlitter between sites, as well as between the microlitter results yielded from the two different methods - pelagic trawl net and gill net, were investigated with T-test. If more than two groups compared, then between group differences were compared with post hoc multiple comparison test. The fish total length (TL) and wet weight (WW) were compared with abundance of ingested litter items using parametric correlation to investigate the effect of fish size to ingested microlitter quantity.

Microlitter particles were recorded in all investigated species and at all sampling sites; however, high variability between the species and among the different sampling areas was observed. Analysis of the occurrence of microlitter in different fish species revealed contamination in 30.5% of the animals sampled. The occurrence of microlitter (FO%) at the regional level was highest in Pärnu Bay (36.0%, n=144), followed by western Estonia (33.1%, n=115), and Gulf of Finland (22.7%, n=191). The highest FO% was measured for P. fluviatilis from the Pärnu Bay (62.3%) and the lowest for P. flesus from Gulf of Finland and C. harengus membras from the Pärnu Bay (15%).

Previous studies have revealed highly different levels of microlitter or microplastic contamination in fish caught in offshore and mid-water trawls or gills. The concentration varies greatly for different studies i.e., 5.5% in the study of Rummel et al. [52], 11% in the study of Lusher et al. [30], 15% in the study of Alomar et al. [53], 23.3% in Giani et al. [54], 29% in Murphy et al. [55], 35% in Boerger et al. [56], and 40-87% in Anastasopoulou et al. [57]. The concentrations of microlitter and plastics found in our study were considerably high when compared to reported concentrations in offshore planktivorous fish herring and sprat collected from the northern Baltic Sea [38] or herring and cod trawled from Polish coast [40]. However, they are considered to be on the same level as what has been found in other species collected in nearshore environments [39, 58] and Caspian Sea [59]. Welden et al. [58] have demonstrated microplastic uptake in 42.5% in plaice M. squinado and in 50% in spider crab P. plastessa. Moreover, it has been shown that benthic fishes often tend to ingest more microlitter than pelagic feeders within the same environment [60, 61], but in this study such difference was not observed.

Our results showed that around 90% of all observed microlitter were fibers and majority of them constituted of plastic and most particles were found at a size class 1 - 5mm or 330µm - 1mm. According to Giani et al. [54] report for the Mediterranean Sea, 66% of microlitter found in two fish species were constituted from fibers, whereas the most common size classes were 0.5 - 1.0mm and <330µm. Abbasi et al. [8] report that in Musa Estuary, Persian Gulf, both pelagic and demersal fish ingested microplastic and recorded mean abundance ranging from 7.8 in tiger prawn to 21.8 in bartail flathead. However, much lower average abundance of microlitter was observed in the current study, where the average in fish amounted to 0.54 items and in bivalves 0.47 items per individual. In general, the majority of all studied individuals had ingested only one item. Sainio et al. [39] reported higher average amount (1.34±0.71) of particles in small coastal fish that had ingested microplastics. However, similarly to small coastal fish investigated in the current study, they also reported that the number of items varied between 1 and 5 particles per fish.

The fact that all studied bivalve species from all sites contained microlitter (31.3% of all analysed individuals) supports the previous findings made by Railo et al. [37] in the Baltic Sea and other earlier studies targeting different geographic regions, suggesting that filter feeding bivalves are efficient in accumulating microlitter [62-64]. Railo et al. [37] have shown that 66% (average 0.4 items/individual) of the blue mussel M. trossulus individuals contained microlitter, with the dominating litter types being fibres (~90% of all microlitter) and plastics made up 8% of all the analysed microlitter particles. In the current study, contamination was observed in 41.7% of M. trossulus (average 0.61 items/individual) individuals and similarly around 90% of all microlitter were fibers; however, contradictory, plastics made up majority of the microlitter.

Despite the population habitat of Harris mud crab R. harrisii in the immediate vicinity to Pärnu harbour in the Pärnu Bay, surprisingly only 22.7% of individuals contained microlitter items. For example, in Celtic Sea nearshore waters, Welden et al. [58] have showed a microplastic uptake of 50% in spider crab P. plastessa. The ingestion of plastic by the Chinese mitten crab Eriocheir sinensis from the Baltic coastal waters (Poland) and the Tagus Estuary (Portugal) revealed 13% individuals from both regions contained microplastic [65].

Due to such extreme variation between locations and target species, one-to-one comparisons between the results presented in the current study and existing studies are limited. To conclude, all investigated species ingest microlitter and can be used as a reference for future studies regarding the use of indicator species for monitoring microlitter pollution in the coast of Estonia. However, neither a target species nor an established threshold values for all the MSFD marine waters are not yet officialised. It is being proposed that suitable bioindicator species which have a wide distribution in the MSFD areas (incl. Baltic Sea) are used, such as anchovy (Engraulis encrasicolus), sprat (Sprattus sprattus), Atlantic horse mackerel (Trachurus trachurus) and Atlantic mackerel (Scomber scombrus) [49]. In this case however, only sprat would be applicable when it comes to the Estonian marine area. Moreover, all these species are pelagic and therefore, other benthic/demersal target species must be chosen at a regional level. As no single species can provide full coverage over all Europe’s marine sectors, a range of species is needed to monitor ingested litter [48]. It is also being suggested that monitoring of microlitter should include sentinel species such as filter feeders, which are known to ingest plastic particles (e.g., blue mussels) [43].

To assess MSFD criterion D10C3 we suggest following specific indicators for Estonian sea area: D10C3.1 - Amount of microlitter ingested by fish; and D10C3.2 - Amount of microlitter ingested by invertebrates. Target species were selected according to criteria set out by Matiddi et al. [49]: a) be representative of specific environmental compartments, b) have a commercial value, c) have a wide distribution in the MSFD (and RSCs) areas, d) already be described as regular litter consumers by different research studies. Selection of target species in the current study was set to represent those found within the entire Estonian sea area and fulfil the criteria on the above-mentioned list. The selected demersal fish species are perch P. fluviatilis and flounder P. flesus sensu lato and pelagic fish species baltic herring C. harengus membras and sprat S. sprattus. Sprat is also being suggested by MSFD Technical Group on Marine Litter [49] and as being widely distributed in most EU countries, and it can be important for assessment and comparison with other, and with more distant, areas. Population structure of Baltic herring is more complex in Estonian marine area, which is constituted of several sub-populations [66] and therefore potentially reflect situation better at the local scale. The selected species of filter - feeding benthic macroinvertebrates constitute the two abundant and widely distributed bivalves in Estonian coastal sea - blue mussel M. trossulus and Baltic macoma M. balthica and omnivorous Harris mud crab R. harrisii.

This study was designed primarily to assess which species would be a good matrix for future status and trends monitoring for microlitter and microplastics, aiming to be a direct and concrete contribution to the implementation of the main legislative marine litter related frameworks in the Baltic Sea region, precisely the EU Marine Strategy Framework Directive [41]. As a result of the current study, baseline values of amount and composition of microlitter in selected species are presented, which, from now on, enables to assess changes (Tables 4 and 5). In the current study insufficient amount of sprat individuals were analysed to calculate the base values and to give adequate estimation of the amount and composition of microlitter in the species. Performing targeted catches and collection of sprat from International Council for the Exploration of the Sea's (ICES) coordinated trawl surveys in the Baltic Sea are necessary to estimate the amount of microlitter in this species. It is suggested, that sampling for analysis of litter in fish should be part of already established surveys [48]. The assessment results are reported by aggregating the percentage of individuals that have ingested microlitter and the average number of microlitter items per individual. These results are then analyzed in relation to the size of the sampled individuals (weight and length) from various sites within the Estonian marine area.

Table 4. National aggregated results of the amount of microlitter (ML) items in relation to size (weight and length) of the individual sampled and proportion (%) of individuals ingested microlitter per selected macrozoobenthic organisms

Note: Proportion (%) of microplastic (MP) out of the total of ML is indicated

Table 5. National aggregated results of the amount of microlitter (ML) items in relation to size (weight and length) of the individual sampled and proportion (%) of individuals ingested ML per selected fish

Note: Proportion (%) of microplastic (MP) out of the total of ML is indicated