Microplastics in the aquatic and terrestrial environment: sources (with a specific focus on personal care products), fate and effects

Sampling

While in some cases bulk water samples were taken (e.g. [18]), volume-reduced sampling methods have generally been employed to sample microplastics from water: neuston nets for the sea surface layer, and zooplankton nets for sub-surface water [6, 11, 19, 20]. Mesh width of the sampling nets was most commonly 300–390 µm [11]. The sampling method has a strong influence on the study results, especially concerning the smallest microplastics, which require a sufficiently small mesh width or bulk sampling, depending on their size [21].

Samples from subtidal sediments are taken with sediment sampling equipment, e.g. grab samplers [6, 11, 22]. For coastal sediments, direct sampling of visually identified microplastics (e.g. by hand or using tweezers) has been used. This method can be employed to sample larger microplastics such as plastic resin pellets (typical diameter: 3.5 mm [23]) from the surface of sandy beaches. However, it is not suitable for sampling microplastics that do not have a characteristic shape, are smaller than plastic resin pellets or are mixed with other debris [6, 11]. In these cases, bulk sampling is required to avoid an underrepresentation of small microplastics. As the distribution of microplastics on beaches is often heterogeneous, attention needs to be paid to sample size, replication and the location of the sampling sites on the beach [8].

Sample processing

To recover microplastics from bulk or volume-reduced water samples, a sieving or filtration step is commonly used. Alternatively, microplastics have been collected from the surface of the water sample using tweezers. Yet, there is a high likelihood of bias when visually collecting microplastics [11, 24]: the lowest particle size in studies, in which sieves were used, was smaller than in studies, in which particles were visually collected from the samples [11].

To separate microplastics from other material recovered by the abovementioned sieving, filtration or density separation procedures, visual sorting has in most cases been used. However, visual differentiation of microplastic particles from other debris or grains of sand is difficult and only suitable for microplastics larger than approx. 1 mm [6, 11, 22, 30]. Transparent or white particles require careful differentiation methods, e.g. examination under high magnification using a fluorescence microscope to confirm the absence of cellular structures [11, 30].

Recently, methods for chemical and enzymatic clean-up have been developed. Treatment of samples with chemicals (e.g. 30–35 % H₂O₂, 20 % HCl, 30 % NaOH), partly combined with increased temperatures, has been used to remove organic material and calcified structures. However, chemical treatment may lead to a partial or complete degradation of microplastics [24, 26, 28, 31, 32, 33]. For this reason, digestion with enzymes (proteinase, cellulase, lipase and chitinase) has been recommended to separate microplastics from organic material [31, 32]. Chemical (e.g. 69 % HNO₃, 10 % KOH, 30 % H₂O₂) and enzymatic (proteinase) methods have also been used to extract microplastics from tissues of aquatic organisms and from the digestive tract content of fish [32, 33, 34, 35, 36, 37, 38, 39].

Identification and quantification

Characteristics such as density and colour have been used to identify the polymer type of microplastics. However, these characteristics can be affected by weathering. As mentioned above, non-plastic microparticles and microfibers may be wrongly classified as microplastics [11, 30, 40]. When analysing microparticles, which had been visually classified as microplastics, with Fourier transform infrared spectroscopy (FT-IR), up to 70 % of the particles were not confirmed to be plastics [11]. Using scanning electron microscopy combined with energy dispersive X-ray spectroscopy (SEM–EDS), Eriksen al. [41] showed that numerous particles, which had been visually identified as microplastics, were aluminium silicates from coal ash and coal fly ash. Consequently, further analyses are required (1) to unequivocally identify microplastics and (2) to obtain information on their resin composition [6, 8, 11, 24, 30, 40, 41]. The use of infrared (IR) and Raman spectroscopy has been highly recommended for this purpose [11, 30]. In EC [6], it is suggested that all particles with a size between 20 and 100 µm and 5–10 % of the particles with a size between 100 µm and 5 mm should be further analysed by IR or Raman spectroscopy. Infrared spectrophotometry and FT-IR are probably the most commonly used method to identify the chemical composition of microplastics. If coupled with microscopy they can be used to identify microplastics with a size >20 µm [6, 11, 30, 31]. Raman spectroscopy combined with microscopy has a higher resolution (approx. 1–2 µm [26, 30, 42]). Several other methods such as pyrolysis–gas chromatography combined with mass spectrometry (Pyr-GC/MS), high temperature gel-permeation chromatography (HT-GPC) with IR detection, SEM–EDS and thermoextraction and -desorption coupled with GC/MS [24, 41, 43, 44, 45] have been developed. With Pyr-GC/MS, both the polymer composition of microplastics and organic additives (e.g. antioxidants and plasticisers) can be analysed simultaneously [24, 43].

The abundance of microplastics is commonly indicated as numerical or mass concentration: (1) for the sea surface layer as number or weight of items per area, (2) for the water column as number or weight of items per volume and (3) for sediments as number or weight of items per sediment area or sediment weight, referring to sediment wet (ww) or dry weight (dw) (see also Additional file 1: Tables S1, S2). Due to the non-standardised units of quantification, the comparison of different studies can be very difficult [11, 29, 46]. It has therefore been suggested that all studies should provide sufficient information to allow converting units, e.g. from items per area to items per volume [11]. Preferably, both numerical and mass concentration should be indicated [8].

Lower size limit of the sampled microplastics

The lower size limit of microplastics sampled in the environment is obviously determined by the sampling and processing methods. Microplastics larger than 300 µm were sampled in most cases from seawater and microplastics larger than 500 µm from sediment [11]. When density separation with a subsequent filtration step is used, smaller particles can be retrieved from sediment. The smallest microplastics sampled from sediment had a diameter of 1 µm. With current state-of-the-art techniques it is most likely not possible to representatively sample and unequivocally identify microplastics with a size below 1–2 µm [11, 26].

Quality control

Given that microplastics research is still in a relatively early stage, inter-laboratory comparisons of protocols for sampling, processing and analysis and certified reference materials are lacking [6, 8, 29, 37]. There is little information on the recovery rates of different sampling and processing methods [26]. Especially small microplastics have a strong tendency to adsorb to surfaces. Therefore, care has to be taken to avoid overlooking particles that adhere to the devices used to collect and process samples [11, 26].

A further important issue is potential contamination of samples by particles originating e.g. from the clothing of workers, the used equipment and the ambient air. As far as possible, such contamination has to be reduced [6, 11, 26, 34, 47]. It has also been suggested that samples should be processed in cleanrooms or cleanroom cabinets and that procedural controls should be included to verify contamination during sample processing [6, 11, 28, 47, 48, 49].

Summary: methods to sample, process and analyse microplastics in the environment

Appropriate sampling, extraction and identification methods are required to representatively sample and unequivocally identify microplastics in the environment. Visual sampling and sorting of small particles (<0.5–1.0 mm) and visual identification of microplastics are not considered reliable, since particles may be overlooked or wrongly classified as microplastics. Quality controls have to be included to verify the efficiency of the used methods and the absence of background contamination.

Primary microplastics

Microplastic particles are used as exfoliants in certain product segments of specific personal care products, such as hand cleaners, facial cleaners and toothpaste [51]. In the US patent for skin cleaners containing plastic microparticles, polyolefin particles with a size of 74–420 µm and an amorphous shape without sharp edges were described as appropriate for use as exfoliants [52]. The used polyolefins include polyethylene (PE), polypropylene (PP) and polystyrene (PS; for abbreviations for the plastic resins see also Table 1). Gouin et al. [53] estimated that in 2012, approx. 6 % of the liquid skin cleaning products marketed in the European Union, Norway and Switzerland contained microplastics. Based on a survey conducted by Cosmetics Europe, PE accounted for 93 % of the microplastics used in skin cleaning products in these countries in 2012. The products typically contained between 0.05 and 12 % of microplastic particles, with the size of most particles ranging from 450 to 800 µm [53]. When analysing skin cleaners, spherical particles (mostly with a rough surface), threads and irregularly shaped particles consisting of PE and PS, and mainly having a blue or white colour were identified [45, 51, 54, 55, 56]. Microplastics are also used in medical applications, e.g. in dentist tooth polish, and as carriers to deliver active pharmaceutical agents [51, 57]. After use, microplastics from personal care products and such medical products can reach the environment via wastewater.

Microplastics are also used in drilling fluids for oil and gas exploration and in industrial abrasives, i.e. for air-blasting to remove paint from metal surfaces and for cleaning different types of engines [55, 57, 58, 59]. Industrial abrasives contain e.g. acrylic, PS, melamine, polyester (PES) and poly allyl diglycol carbonate microplastics [41]. If not used in closed systems and disposed properly, they can end up in the environment [57].

Raw materials used for the fabrication of plastic products (pre-production plastics), namely plastic resin pellets or flakes and plastic powder or fluff, are another important source of primary microplastics. They can reach the environment after accidental loss during transport or with run-off from processing facilities, i.e. often as a result of improper handling. Similarly, residues from plastic processing factories and regranulate produced during plastic recycling can end up in the environment [9, 25, 58, 59, 60, 61, 62]. Concentrations of plastic resin pellets in the environment were high from the 1970s to the 1990s [63, 64]. Highest concentrations of pre-production pellets (up to 100,000 pellets/m of beach) were often found on beaches close to plastic producing or processing sites [29, 60, 61]. In subsequent years, concentrations of pre-production plastics in the environment generally declined, probably due to improved practice during handling [63, 65, 66]. Still, high concentrations have been found in some studies close to production facilities ([67, 68, 69]; see also section “Occurrence of microplastics in the aquatic and terrestrial environment”).

Macroplastics as sources of secondary microplastics

Since secondary microplastics are generated when larger plastic materials fragment, sources of macroplastics and their routes of entry into the environment are considered in this section. It has been estimated that about 75–90 % of the plastic debris in the marine environment originates from land-based and about 10–25 % from ocean-based sources [25, 70]. General littering, dumping of plastic waste and loss from inappropriately managed landfill sites and during waste collection are assumed to be the most important routes of entry of plastic materials into the environment. Windblown litter is also lost from recycling facilities [60, 70, 71, 72]. In this context, it should be noted that a large percentage of the produced plastics is used for packaging, i.e. for products with a short service life [1]. In industrialised countries, waste that is deposited in landfills is usually covered regularly with soil or a synthetic material, and the landfill is surrounded by a fence to prevent that debris is blown away. However, in developing regions this is often not the case [71, 72, 73]. In addition, large amounts of plastic debris can enter the marine environment during natural disasters such as hurricanes, tsunamis and strong sea [18, 74, 75].

Low-density polyethylene (LDPE) films, which are used in large volumes to protect agricultural crops, suppress weeds, increase temperature and retain irrigation water in the soil (‘plastic mulching’), are a further relevant source of microplastics in the environment. If these thin plastic foils embrittle, the fragments can end up in the soil [72, 76, 77]. Synthetic polymer particles, such as expanded PS flakes with a size of approx. 5–15 mm (Styromull^®) and polyurethane (PU) foam, are also used in horticulture to improve soil quality and as composting additive [78, 79].

Moreover, synthetic textiles are an important source of microplastics. In 2013, 54.4 million t of synthetic fibres were produced worldwide with PES (44.6 million t) being the dominant fibre type [80]. Browne et al. [81] quantified the number of fibres released when washing PES blankets, fleeces and shirts in domestic washing machines. The washing machine effluent contained approx. 120 (blanket) to 300 fibres (fleece) per L. Overall, >1900 fibres were given off from the evaluated PES fleece during a single wash. Synthetic textile fibres are also released to air and dust, either during normal use [57] or during tumble drying [77]. In addition, synthetic fibres are released from hygiene products, e.g. if improperly disposed into wastewater [31].

A number of other sources of microplastics in the environment have been identified. Abrasion from car tyres has been considered as very relevant [46, 57]. In addition, many ship paints and other protective paints contain synthetic polymers, e.g. alkyds, poly(acrylate/styrene), PU and epoxy resins. Microplastics may be released by spills during application of the paint, by abrasion during use of the painted product and during paint removal [21, 57]. Microplastics are also released as a consequence of abrasion from other plastic materials such as household plastics [31, 57].

Ocean-based sources of marine litter include material lost or discarded from fishing vessels, aquaculture facilities, merchant ships, recreational boats, offshore oil or gas platforms and during military activities. Cargo lost from merchant ships may lead to a significant input of plastics into the marine environment [25, 58, 60, 72, 82]. Although dumping of plastic wastes at sea is prohibited since 1988, there are indications that plastic waste from a considerable number of vessels has still been dumped at sea—mainly due to economic reasons [46, 58, 83].

Marine environment

Triggered by the detection of unexpectedly high levels of microplastics in oceanic convergence zones (e.g. [3, 25, 119]), the occurrence of microplastics has mainly been investigated in the marine environment (including shorelines). Microplastics have been found in the oceans worldwide [71, 120], including remote regions such as Antarctica and the deep sea [58, 121, 122, 123, 124]. The distribution of microplastics in the oceans is very heterogeneous. High concentrations have been found close to industrial centres and metropolitan areas [22, 58, 110, 125], in enclosed or semi-enclosed seas such as the Caribbean and the Mediterranean Sea and in gyres [8, 64, 71, 126].

Clear increases in the abundance of microplastics in seawater have been found over larger temporal scales. Thompson et al. [127] evaluated archived plankton samples collected with continuous plankton recorders in the northeast Atlantic at 10 m depth. Samples were examined microscopically and unusual fragments were identified by FT-IR. The concentration of microplastics significantly increased from the 1960s and 1970s (approx. 0.01 items/m³) to the 1980s and 1990s (approx. 0.04–0.05 items/m³). Based on published data and own investigations using a manta trawl and visual identification of microplastics Goldstein et al. [128] evaluated temporal changes in microplastic levels in the surface layer of the North Pacific gyre. They found that between 1972–1987 and 1999–2010 the median numerical concentration of microplastics had increased by a factor of approx. 140 (from 0.003 to 0.425 items/m³) and the median mass concentration by a factor of 1000 (from 0.003 to 3 mg/m³). Law et al. [20] re-analysed the temporal trend using a combined dataset including the data of Goldstein et al. [128] and their own data obtained by sampling using neuston nets and visual identification of small plastic items (typically mm-sized; see Additional file 1: Table S1). To reduce the bias, which is caused by an increased sampling frequency in areas with high plastic levels in recent years, they calculated average concentrations of small plastic items for each 1° × 1° (latitude × longitude) area. On the basis of these data, Law et al. [20] concluded that both mean and median numbers of small plastic items in the surface layer of the North Pacific gyre increased by a factor of approx. 10 between 1972–1985 and 2002–2012. It is controversially discussed if the increasing trend has continued since the 1990s [20, 65, 71, 128, 129]. Due to large spatial and temporal variability in the concentrations of microplastics, it is difficult to detect smaller increases [19]. High variability of the levels of small plastic items at a smaller scale (i.e. within tens of km) is caused by local wind-driven turbulences and local circulation patterns [20]. Higher concentrations in the sea surface layer are e.g. measured during low-wind conditions [19, 126]. Temporal variability can be high, e.g. due to variations in oceanic circulations related to El Niño events [71, 99]. Due to the embrittlement and fragmentation of larger plastic items present in the oceans, it has been predicted that the overall abundance of microplastics will increase in the future [25, 71, 99, 120].

In the following, an overview is provided of the concentration ranges of microplastics in the marine environment (including estuaries) that is based on the review of Hidalgo-Ruz et al. [11] and selected recent publications. The compiled data shall primarily serve as background information when evaluating the environmental relevance of test concentrations in studies on uptake and effects of microplastics. For this reason, main focus is placed on studies reporting numerical concentrations per water volume, sediment weight or sediment volume. As outlined above, most data on the occurrence of microplastics in the marine environment have been obtained using visual selection and often also visual identification of microplastics [11]; the results of quality controls are only reported in some cases (e.g. [49, 110, 123]). Although data obtained by visually selecting and/or identifying small microplastics should be considered with care, they provide a first impression on the abundance of microplastics in marine systems and were, thus, not excluded from the present review. However, an attempt was made to especially include recent publications using state-of-the-art sampling, extraction and identification techniques (see Additional file 1: Table S1 for information on the used methods).

As most consumer plastics are at least initially buoyant in seawater, the abundance of microplastics in the sea surface layer, the upper approx. 20 cm of the water column, was addressed in many studies. Buoyant micro- and macroplastics have been found to accumulate in the North Atlantic, South Atlantic, North Pacific, South Pacific and Indian Ocean gyre [65, 119, 128, 129, 130]. In the North Pacific gyre, mean abundance of visually identified plastic items (0.33 items/m², mostly fragments, plastic films and fibres with a size <5 mm) in 3 out of 11 samples was found to be higher than plankton abundance [119]. Based on the review of Hidalgo-Ruz et al. [11] and the evaluated recently published studies (Additional file 1: Table S1), concentrations of microplastics in the sea surface layer range from 0 to 12.3 items/m² [20] and from 0 to approx. 8700 items/m³ (i.e. 8.7 items/L [11]). A much higher abundance of synthetic polymer particles was found in the sea surface microlayer, i.e. the upper 1 mm of the water column (Additional file 1: Table S1). Mean abundance was 195 items/L for paint particles, mainly alkyd and poly(acrylate/styrene) polymers and 16 items/L for non-paint microplastics [21].

In the water column below the sea surface layer, concentrations of microplastics can be expected to be lower than at the water surface. According to the evaluation of Hidalgo-Ruz et al. [11], concentrations of microplastics in the water column ranged from 0.014 to 12.5 items/m³. Considerably higher concentrations were detected in a recent study [18] in the north-eastern Pacific and close to the coast of British Columbia (Canada). Concentrations of microplastics in seawater sampled from a depth of 4.5 m ranged from 8 to 9180 items/m³. They were lowest at offshore sites in the north-eastern Pacific (mean: 279 items/m³) and higher at nearshore sites (1710–7630 items/m³). Approx. 75 % of the microplastics were fibres; the percentage of fibres increased near the shores. As noted by the authors, fibre content was underestimated: brightly coloured brittle fibres were observed in most samples, but could not be quantified since they were eliminated during acid digestion. High concentrations of microplastics (0.5 mm to approx. 5 mm) were also found in the Yangtze estuary in China [131]. In water samples collected at 1 m depth, the average concentration of microplastics was 4137 items/m³.

Coastal and, especially, subtidal sediments appear to be sinks for microplastics [89, 115, 123]. As pointed out by Hidalgo-Ruz et al. [11] microplastic concentrations in sediments tend to be much higher than those in the sea surface layer and in the water column. Microplastics have been detected in coastal sediments (in most cases beaches) around the world [81, 99]. At some sites, they accounted for 80 % of the intertidal plastic debris [132].

Microplastic levels vary greatly between beaches. Claessens et al. [110] and Van Cauwenberghe et al. [105] evaluated microplastics (≤1 mm) on Belgian beaches. Claessens et al. [110] recorded an average concentration of 93 items/kg sediment dw with plastic fibres accounting for 88 % of the microplastics, granules for 7 % and plastic films for 5 %. Van Cauwenberghe et al. [105] found a mean concentration of 13 items/kg sediment dw. In both studies, highest microplastic levels were observed at the high tide line, where resuspension of particles during flooding occurs less frequently. In sediments from a beach on the East Frisian Island Norderney (Germany), levels of microplastic particles (<1 mm) were low (mean values: 1.7, 1.3 and 2.3 particles/kg sediment dw). Due to background contamination with fibres, fibre content of the sediment was not evaluated [49].

Much higher concentrations of microplastics (at least 2 dimensions <5 mm) were found on beaches of three Canary Islands [133]. They ranged from 1 to 30 (Fuerteventura), <1–109 (Lanzarote) and <1–90 g/L sediment (La Graciosa). Low microplastic levels were found on rocky shores and on beaches being completely submerged at high tide (i.e. offering no space for deposition of plastics). Highest levels were recorded on the northern coasts, which are reached by currents likely to transport plastic debris. Carson et al. [134] detected partly extremely high levels of small plastic particles (<10 mm) on southern Hawaii, on two beaches located close to the North Pacific subtropical convergent zone. Kamilo beach sediment contained on average 1.3 % (w/w) of small plastic particles. More than 50 % of the particles were found in the upper 5 cm of the sediment, which contained on average 3.3 % (w/w) of small plastics (the maximum value was 30 %). At Waikapuna beach, mean plastic content was 0.03 %. Again, most plastic occurred in the top 5 cm of the beach, where plastic content was 0.1 %. The plastic particles analysed by FT-IR mainly consisted of PE (85 %) and PP (14 %). Elevated concentrations of small plastic particles (mean: 805 items/m²) were also found on beaches on the Easter Islands [101]. Lee et al. [102] recorded high microplastic levels (mean values: 8205 items/m² before and 27,606 items/m² after the rainy season) on South Korean beaches close to the estuary of the Nakdong River, which flows through a densely populated metropolitan area.

Microplastic levels in subtidal sediments from harbours, coastal and offshore areas of the Belgian continental shelf were investigated by Claessens et al. [110]. Harbour sediments contained significantly more microplastics (mean: 167 items/kg sediment dw) than coastal (92 items/kg sediment dw) and offshore sediments (105 items/kg sediment dw). In harbour sediments, fibres accounted for 40 % of the microplastics, granules for 34 %, plastic films for 4 % and PS spheres for 22 %. Microplastics in coastal and offshore sediments consisted of 68 % fibres, 30 % granules and 2 % plastic films. High concentrations in harbour sediments were thought to be related to local input and to the fact that the studied harbour areas were partly enclosed. The analysed fibres consisted of PA, polyvinyl alcohol and PP, the granules of PS, PE and PP. Concentrations of microplastics (<1 mm) in sediments sampled at approx. 1 m depth in the Lagoon of Venice (Italy) were studied by Vianello et al. [22]. Levels of microplastics ranged from 672 to 2175 items/kg sediment dw. They were higher in the inner area of the lagoon, where water currents were low (i.e. more sedimentation occurred). Most microplastics consisted of PE (48 %) or PP (34 %). Irregularly shaped fragments accounted for 86 % of all microplastics, fibres for 11 %, films for 2 % and pellets/granules for 1 %.

In an early study using polarised light microscopy and semi-quantitative evaluation, Habib et al. [92] found that abundance and size of textile fibres in subtidal sediments decreased with increasing distance from a WWTP. Browne et al. [81] sampled subtidal sediments from two sites, where sewage sludge had been disposed, and two reference sites (Additional file 1: Table S1). Although disposal of sewage sludge had been stopped more than 10 years before, the former disposal sites contained 4 times (North Sea) and 2.5 times (English Channel) more microplastics than the reference site.

Microplastics were also detected in deep sea sediments, e.g. in the northeast North Atlantic and southwest Indian Ocean and in the Mediterranean Sea at water depths between 300 and 3500 m [123]. The sediments contained 28–800 microplastics/L. Notably, exclusively fibres were found. Likewise, fibres were the dominant type of microplastics found in sediments from the Kuril-Kamchatka Trench (northwest Pacific) at water depths of 4869–5766 m. In addition, paint chips and fragments were recorded. The microplastics did not only occur in the upper 2 cm of the sediment, but also in deeper sediment layers [124].

Freshwater environment

So far, there are only relatively few studies on microplastics in the freshwater environment. These studies have focused on larger rivers and lakes, while there are no data on the occurrence of microplastics in smaller streams or lakes [72, 95, 111].

During a 2-year survey, Lechner et al. [68] evaluated the abundance of small plastic items (0.5–20 mm) in the surface layer of River Danube between Vienna (Austria) and Bratislava (Czech Republic). In 2010, mean concentration of small plastic items was 0.938 items/m³; 86 % of these items were pre-production plastics. In 2012, plastic abundance was much lower (0.055 items/m³); pre-production plastics accounted for 31 % of these items. The high levels of pre-production plastics in 2010 were apparently caused by leakages in the pipe system of a plastic producer and by a strong rain event, during which pellets were washed into the Danube [135, 136, 137]. Recently, a similar mean microplastic concentration (0.29 items/m³) was derived for the surface layer of the River Rhône [138]. The microplastics mainly consisted of fragments (40 %), foams (37 %) and fibres (14 %). A clear effect of a municipal WWTP on microplastic levels was demonstrated in North Shore Channel (Chicago, USA). Mean microplastic concentration in the surface layer increased from 1.94 items/m³ upstream of the WWTP to 17.93 items/m³ downstream of the WWTP [139]. In sediments from St. Lawrence River (Canada), plastic granules with diameters of 0.4–2.2 mm and mainly grey or black colour were detected. Based on their melting point, it was assumed that the granules consist of PE. Personal care products were mentioned as possible source. However, highest concentrations (136,926 items/m²) were found in the effluent canal of a nuclear power plant, while concentrations at the other nine sampling sites ranged from 0 to 243 items/m² [140]. This suggests that the granules may originate from a source other than personal care products that remains to be identified.

In a preliminary study of microplastic (0.3–5 mm) levels in the surface layer of Lake Geneva (Switzerland), a microplastic concentration of 0.048 items/m² was derived [141]. A similar mean microplastic level (0.091 items/m²) was determined in six lakes located (or partly located) in Switzerland [138]. Fragments, foams, films and fibres were dominant. Microplastics (0.355 to approx. 5 mm) were also analysed in three of the Great Lakes: Lake Superior, Lake Huron and Lake Erie [41]. The mean concentration in the surface layer of these lakes was 0.043 items/m²; 81 % of all microplastics had a size ≤1 mm. The highest microplastics levels (0.463 items/m²) were detected in Lake Erie, downstream of the cities of Detroit, Cleveland and Erie. Fragments, pellets and foams were found most frequently. The particles also included green, blue and purple spheres, which had a similar size, shape, colour and elemental composition as microbeads from facial cleansers analysed by Eriksen et al. [41]. However, more specific information on the abundance of the coloured spheres in the surface layer from the Great Lakes is not provided. Only slightly lower levels of microplastics (0.355 mm to approx. 5 mm) were recorded in the surface layer of Lake Hovsgol, a large mountain lake in northern Mongolia [103]. Fragments, films and lines/fibres were the dominant items. The average microplastic concentration was 0.0203 items/m². Around the lake, there is no industry, and population density is low. Free et al. [103] assumed that the complete lack of a waste management system (waste is burned, buried or dumped) and of wastewater treatment are the causes for the relatively high microplastic levels. Microplastic concentrations were highest in the most populated and most touristic part of the lake. A parallel survey of macroplastic showed that household plastics (plastic bottles and bags) and fishing gear were the dominant macroplastic items at the lake shores.

In sediments sampled at the shores of the rivers Rhine and Main in Germany, microplastic concentrations of 228–3763 items/kg dw (River Rhine) and 786–1368 items/kg dw (River Main) were recorded with PE, PP and PS being the dominant polymers [112].

Imhof et al. [142] evaluated levels of microplastics (<5 mm) in sediment samples from two beaches of Lake Garda (Italy). Microplastic concentration was much higher at the northern (1108 items/m²) than at the southern lake shore (108 items/m²). This difference was attributed to the prevailing wind direction and the resulting water circulation. PS, PE and PP particles were most frequently found. With regard to the smallest microplastics (<500 µm), PA and polyvinyl chloride (PVC) were also relevant. Most particles were classified as fragments; signs of degradation/fragmentation were identified by SEM. At beaches of six lakes in Switzerland, microplastic levels ranged from 20 to 7200 items/m². Foam, fragments and fibres were found most frequently [138].

Terrestrial environment

Although microplastics obviously enter the terrestrial environment (e.g. due to littering and application of sewage sludge to land) and soils have been assumed to be a sink for microplastics, there are only extremely few data on microplastic concentrations in the terrestrial environment [71, 72, 77]. It was suggested to use synthetic fibres as indicators of previous sludge application to land [92, 93]. Using polarised light microscopy, Zubris and Richards [93] showed that up to 15 years after sludge application, levels of plastic fibres in soil of a long-term experimental field site clearly exceeded levels at control sites.

Summary: occurrence of microplastics in the aquatic and terrestrial environment

As a consequence of their persistence and long-range transport, microplastics are ubiquitous in the marine environment. Notably, beaches (where extremely high levels have been recorded at some cases) and subtidal sediments appear to be sinks for microplastics. Based on first investigations, microplastic levels in large rivers and lakes are similar to those in the marine environment. The occurrence of microplastics in smaller streams and lakes and, especially, in the terrestrial environment remains to be investigated.

In aquatic environments, fibres, fragments, granules, films and styrofoam particles are the most commonly found particle types, while PE, PP and PS are the most frequently found polymers. Due to the lack of distinct characteristics, it is generally not feasible to unequivocally identify personal care products as source of microplastics detected in the environment.

Uptake by aquatic organisms

In laboratory experiments, ingestion of microplastics has been demonstrated for a number of marine invertebrates including ciliates, cnidarians, rotifers, annelids, copepods, cladocerans, amphipods, mysids, euphausiids, barnacles, mussels and tunicates [8, 127, 145, 146, 147] and fish [148]. For example, Lee et al. [149] investigated the uptake of nano- (50 nm) and microsized (0.5 and 6 µm) fluorescently labelled PS spheres by the copepod Tigriopus japonicus. Copepods were exposed for 24 h to concentrations of 9.1 × 10¹⁴ (50 nm), 9.1 × 10¹¹ (0.5 µm) and 5.3 × 10⁸ items/L (6 µm). Spheres of all three sizes were detected in the gut of the copepods. In the marine amphipod Allorchestes compressa, which had been exposed for 72 h to a very high concentration (100 g/L) of PE microplastics (11–700 µm), on average 19 particles per amphipod were detected [150]. Ingestion of microplastics was also demonstrated in larvae of the sea urchin Tripneustes gratilla held for 5 d at concentrations of 10³–3 × 10⁵ items/L of fluorescent PE microspheres (10–45 µm). At the highest concentration up to 31 % of sea urchin larvae contained microplastics in their stomachs, at the lowest concentration up to 5 % (on average 1–2 microspheres per larva; [151]).

In addition to being ingested, small microplastics might also be taken up via the gills [8]. For mussels, it has been suggested that uptake via the gill surface (endocytosis) may be relevant for smaller microplastics, while larger particles are taken up via the digestive system ([152], see next section for further details on this study). In shore crabs (Carcinus maenas) exposed to PS microspheres (8–10 µm), exclusively via the ventilation route, microspheres were detected on the gill surface but not in the gill tissue [153].

Uptake of microplastics by freshwater organisms has so far only been addressed in relatively few studies. As to be expected, the available data show that microplastics are also ingested by freshwater organisms. Fluorescent PS microparticles (1 µm) were ingested by protozoans (Paramecium sp.) and Daphnia sp. [144]. Nano- (20 nm) and microsized (1 µm) fluorescent carboxylated PS spheres were taken up by neonates and adults of Daphnia magna [154]. Ground fluorescent polymethyl methacrylate particles (approx. 30 µm) were ingested by D. magna, the ostracod Notodromas monacha, the amphipod Gammarus pulex, the snail Potamopyrgus antipodarum and the oligochaete Lumbriculus variegatus [142].

Transfer from the intestinal tract to the surrounding tissue or circulatory system

After ingestion, microplastics can remain in the digestive tract, be excreted or absorbed from the digestive tract into the body tissue [132, 145]. Lugworms (Arenicola marina) exposed to sediment containing pre-production PS particles (400–1300 µm, 7.4 % of sediment dw) ingested these microplastics. However, no translocation of the relatively large PS particles from the gut to the tissue was recorded [155]. Hämer et al. [156] fed marine isopods (Idotea emarginata) with food containing fluorescent PS microspheres (10 µm), PS fragments (1–100 µm) or acrylic fibres (0.02–2.5 mm). Microplastics were detected in stomach and intestine, but not in the midgut where nutrients are resorbed. Passage of the microplastics to the midgut was most likely impeded by filter structures in the isopods’ proventriculus. In D. magna, fluorescent-carboxylated PS nano- and microspheres (20 nm and 1 µm diameter) were mainly observed in the gastrointestinal tract, but also in structures assumed to be oil storage droplets. It was concluded that the PS spheres are able to cross the gut epithelium [154]. In shore crabs (C. maenas), which were fed with mussels (Mytilus edulis) pre-exposed to PS microspheres, translocation from the intestinal tract to haemolymph, hepatopancreas, ovary and gills was demonstrated for microspheres with 0.5 µm diameter [157]. By contrast, larger microspheres (8–10 µm diameter) were only detected in the intestine but not in the haemolymph of shore crabs [153].

Browne et al. [145] kept mussels (M. edulis) for 3 h in a suspension of fluorescent PS microspheres (3.0 and 9.6 µm; 4.3 × 10⁴ items/L). Microspheres were detected in the haemolymph and inside the haemocytes. The smaller microspheres occurred in significantly higher abundance in the haemolymph than the larger ones. Von Moos et al. [152] exposed mussels for 3–96 h to 2.5 g/L of HDPE fluff consisting of non-uniformly shaped particles with a size between 0 and 80 µm. The concentration of HDPE fluff corresponds to approx. 2.7 × 10⁷ to 3.6 × 10⁷ items/L (NR von Moos, personal communication). HDPE microparticles were detected on the gill surface and in blood lacunae of the gills, as well as in the intestine, digestive gland and connective tissue.

In a very recent study, mullets (Mugil cephalus) were held for 7 days in water containing 33.8 mg/L of PE or PS particles with a size of 0.1–1 mm (nearly 2500 particles/L [33]). Microplastics were not only found in the gastrointestinal tract (approx. 10 PE particles and 90 PS particles per fish), but also in the liver of the fish (approx. 1–2 particles per fish for both PE and PS).

Thus, based on the results of laboratory experiments, translocation from the intestinal tract to the circulatory system or surrounding tissue depends on the size of the microplastics with an upper size limit for translocation that appears to be specific for the species or taxonomic group.

Excretion by aquatic organisms

In laboratory experiments, microplastics, which had only entered the organisms’ intestinal tract, were generally excreted within hours or few days. Rapid excretion of microplastics was e.g. reported for the copepod Eurytemora affinis [147]. After 3 h exposure to fluorescent PE spheres (10 µm; 2 × 10⁶ items/L), 67 % of the copepods had ingested microspheres. Following a 12 h post-exposure, only 4 % of the copepods still contained microspheres. In marine amphipods (A. compressa), most ingested PE microplastics were excreted within 2 d [150]. Sea urchin larvae (T. gratilla) egested PE microspheres (10–45 µm) within 7 h [151].

In cases, where microplastics had translocated to the circulatory system and/or the surrounding tissues, excretion was also demonstrated, but was slower. In shore crabs (C. maenas), concentrations of fluorescent PS microspheres (0.5 µm) in the haemolymph decreased from 24 h to 21 days post-exposure. Yet, on day 21 the haemolymph still contained a few microspheres [157]. In mussels (M. edulis), fluorescent PS microspheres (3.0 and 9.6 µm) were—despite excretion with the faeces—still detected in the haemolymph 48 d post-exposure [144].

Trophic transfer

Several laboratory studies have demonstrated that microplastics are transferred in the food chain, i.e. from prey to predator. Trophic transfer of fluorescent PS microspheres (10 µm) from zooplankton to the mysid shrimp Mysis relicta was observed by Setälä et al. [147]. M. relicta, which had been kept together with zooplankton (copepods and polychaete larvae) pre-exposed for 12 h to microplastics, contained PS microspheres in their stomach. Trophic transfer of microplastics from mussels (M. edulis) to shore crabs (C. maenas) was shown for fluorescently labelled PS microspheres with 0.5 µm [157] and 8–10 µm diameter [153]. So far, it is not known if microplastics biomagnify, i.e. if concentrations in the organisms increase at higher trophic level [158].

Field studies with aquatic organisms

The uptake of microplastics has also been investigated in a number of field studies, mainly with marine organisms. Note that in these studies no differentiation between direct ingestion of microplastics and uptake with the food (i.e. trophic transfer) is possible.

Desforges et al. [36] analysed microplastics in two zooplankton species, the copepod Neocalanus cristatus and the euphausiid Euphausia pacifica, sampled in the Northeast Pacific. On average, 0.03 microplastic items/individual were found in N. cristatus, 0.06 in E. pacifica. Goldstein and Goodwin [159] evaluated plastic ingestion by gooseneck barnacles (Lepas anatifera, Lepas pacifica, Lepas sp.), rafting organisms attached to floating substrates at the sea surface. Gooseneck barnacles were sampled from floating debris in the North Pacific gyre; microplastics in their stomachs and intestines were identified visually. Due to the possibility of airborne contamination, fine microfibers were not considered. A subset of microplastic particles was analysed by Raman spectrometry. Microplastics were found in 34 % of the barnacles, with most barnacles having ≤5 particles in their intestinal tract. Nearly all (99 %) particles were degraded fragments (median diameter: 1.4 mm). No blockage of the stomach or intestine was recorded. Analysis by Raman spectrometry revealed that 58 % of the particles consisted of PE, 5 % of PP and 1 % of PS (polymer type of the remaining particles could not be analysed).

Van Cauwenberghe and Janssen [38] investigated levels of microplastics in cultured mussels (M. edulis) reared in the North Sea and Pacific oyster (Crassostrea gigas) reared in the Atlantic. Mussels were either analysed directly for their microplastic content or kept for 3 days in filtered artificial seawater to allow clearing of the gut. Soft tissues were digested in concentrated nitric acid, boiled, diluted with deionised water and filtered (5 µm). Microplastics were identified microscopically, and a subset was analysed by micro-Raman-spectrometry. Samples were processed in a laminar flow cabinet; procedural controls were free of microplastics. In M. edulis, 0.36 microparticles per g soft tissue ww were detected without depuration. After the depuration period, 0.24 items/g soft tissue ww were found. In C. gigas, 0.47 items/g soft tissue ww were found before and 0.35 items/g tissue ww after depuration. As discussed by Van Cauwenberghe and Janssen [38] tissue digestion by nitric acid and boiling may have led to destruction of smaller microplastics and, hence, to an underestimation of microplastic levels in the mussels. With micro-Raman-spectrometry, no unambiguous identification of the particles as microplastic was possible. The authors assumed that this might be due to effects of the digestion procedure on the plastic matrix. Similar microplastic concentrations were reported for M. edulis collected at the French, Belgian and Dutch North Sea coast [122], and slightly lower levels for Mediterranean mussels (Mytilus galloprovincialis) sampled in the estuaries of the rivers Tagus (Portugal) and Po (Italy) and in the Ebro delta (Spain [37]). A. marina from the French and Belgian North Sea coast contained on average 1.2 microplastic particles/g tissue [122].

Plastic concentrations in the gut of small (up to 10 cm standard length) pelagic fish in the North Pacific gyre were studied by Boerger et al. [160]. Fish from six species were caught with manta trawls. Approx. 35 % of the fish had plastic particles (on average 2 items) in their stomach. Most of the ingested plastics were fragments (94 %) and had a size of 1.0–2.8 mm. Lusher et al. [161] evaluated the plastic content in the gastrointestinal tract of 5 pelagic and 5 demersal fish species caught in the English Channel. All items, which were not classified as natural prey of the fish, were analysed by FT-IR. Items consisting of plastic or the semi-synthetic fibre rayon were found in 37 % of the analysed fish (on average 1.9 items/fish). There was no significant difference in the number of ingested items between pelagic and demersal species. The ingested material mainly consisted of fibres (68 %), fragments (16 %) and beads (12 %); 92 % of the items were smaller than 5 mm, with particles of 1.0–2.0 mm size being most common. More than half of the items (58 %) were identified as rayon, 36 % as PA, 5 % as PES, 1 % as PS, 0.3 % as PE and 0.3 % as acrylic. Polymers with a lower density were predominantly detected in pelagic fish species, those with a higher density in demersal species. Microplastic particles (0.2–5 mm; due to the possibility of airborne contamination fibres were not included in the evaluation) were found in the intestinal tracts of 5 out of 7 fish species sampled in the North Sea. Overall, 2.6 % of the fish had ingested microplastics [34]. In an analysis of the microplastic content in the gastrointestinal tract of fish from five commercially used species sampled in the Adriatic Sea, microplastic items were found in 28 % of the fish (1–2 items/fish [33]).

Plastic items, including microplastics, have also been shown to be ingested by seabirds and marine mammals. For instance, plastic content in the stomach of dead, beach-washed northern fulmars (Fulmarus glacialis) is monitored in the context of the OSPAR ecological quality objective for marine litter stating that less than 10 % of northern fulmars should have >0.1 g plastic particles in their stomachs [162]. Fulmar stomachs were shown to contain macro- and microplastics. However, the latter were not quantified separately [163, 164]. Microplastics and small macroplastics (0.5–30 mm) were detected in the scats (faeces) of fur seals (Arctocephalus tropicalis, Arctocephalus gazella) on Macquarie Islands, between New Zealand and Antarctica [165]. Plastic particles were collected visually and analysed by SEM and FT-IR. Most scats contained 1 plastic item; particles mainly consisted of PE and had a size between 2 and 5 mm. Microplastics were also found in the digestive tract of a whale (Mesoplodon mirus) stranded on the Irish coast [166].

Little information is available on the uptake of microplastics by freshwater organisms in the field. In a preliminary study, Sanchez et al. [167] examined wild gudgeons (Gobio gobio), caught in French streams. The digestive tract of the fish was analysed visually for microplastics. Based on first results, 12 % of the gudgeons had microplastics in their intestine. The abundance of microplastics increased with the anthropogenic impact on the rivers. While no microplastics were found in the intestine of fish captured at three sites characterised by a low anthropogenic influence, the highest percentage of fish with microplastics in their intestine (approx. 20–30 %) was recorded at two sites in urban areas.

Uptake and excretion by terrestrial organisms

Very little information is available on the uptake of microplastics by terrestrial organisms. Ugolini et al. [168] performed a laboratory study with the sand hopper Talitrus saltator, an amphipod inhabiting sandy coasts. In T. saltator, which had been fed with dry fish food mixed with 10 % (w/w) of PE microspheres (10–45 µm), ingestion of the microspheres was demonstrated. Within 24 h most microspheres were excreted, within a week all microspheres.

Summary: uptake and trophic transfer

A wide variety of aquatic organisms have been shown to ingest microplastics. In most cases, particles were only detected in the intestinal tract and excreted rapidly. Translocation from the intestinal tract to the circulatory system or surrounding tissues was observed in some species for very small microplastics. In field studies, microplastics were ingested, but the ingested quantities were low. Microplastics are transferred in the food chain. Yet so far, there are no data demonstrating their bioaccumulation or biomagnification.

Physical effects of microplastics

Macroplastics physically affect marine organisms. Especially for air-breathing animals, entanglement may result in death. The ingestion of macroplastic items may reduce the amount of consumed food and, consequently, the organisms’ fitness. Macroplastics can also block the intestinal tract and cause internal injuries [58, 75, 169, 170, 171]. It has been assumed that microplastics cause similar effects in smaller organisms, mainly with regard to the physical obstruction of feeding and digestion [6, 7, 71, 75]. Sharp-edged microplastics may injure gill tissues and the intestinal tract [7, 144]. In the following, the available data on physical effects of microplastics on aquatic organisms are summarised (see also Additional file 1: Table S3). While studies dealing exclusively with effects of nanoplastics were not considered, results of comparative studies of nano- and microplastics have been included.

Microplastics as vectors for pollutants

Microplastics as substrates for organisms

Many marine organisms live attached to debris [190, 191]. Due to plastics, the amount of floating debris in the oceans has greatly increased [190]. This is most relevant in the open ocean, where only very limited floating substrate is available [128, 192]. Plastic debris is often colonised by microorganisms and—depending on its size—also by larger organisms. PE microplastics and small macroplastics collected in the surface layer of the North Atlantic were colonised by a variety of organisms including bacteria, cyanobacteria, diatoms, ciliates and radiolaria [191]. Bryozoans were identified on 3–5 mm-sized microplastics sampled in the surface layer at different sites around Australia [193]. Microplastic particles are also used as oviposition sites by the sea skater Halobates, an insect living at the sea/air interface in the open ocean [128, 192]. The strong increase in microplastics in the North Pacific gyre, which was observed between 1972–1987 and 1999–2010, was associated with a considerable increase in the number of eggs, juveniles and adults of Halobates sericeus. Possible effects of the increased abundance of H. sericeus on other species have so far not been investigated [128].

Given that plastics can be transported over long distances, they may contribute to the dispersal of species [75, 190]. This includes invasive species [190] and species causing harmful algal blooms [194]. Due to their low size, microplastics may especially facilitate transport of microorganisms (including pathogens) and other very small organism. So far, it is not known if the transport of species with microplastics has any significant effect on species assemblages [158, 195].

Effects of microplastics on sediment properties

Carson et al. [134] used artificially constructed beach sediment cores containing 1.5, 7.3, 15.9 and 29.4 % (w/w) of small plastic particles (<10 mm) to investigate effects on water permeability and heat transfer. Notably, mean size of plastic particles was higher than mean grain size of the sediment. At 15.9 and 29.4 % (w/w) plastics, water permeability was significantly increased. With increasing plastic content, sediments warmed more slowly. A 16 % decrease in heat transfer was recorded at the highest plastic concentration. Possible implications of such effects on physical properties of sediments remain to be investigated.