Ingestion and Toxicity of Polystyrene Microplastics in Freshwater Bivalves

Abstract

The ubiquity of microplastics in aquatic ecosystems has raised concerns over their interaction with biota. However, microplastics research on freshwater species, especially molluscs, is still scarce. We, therefore, investigated the factors affecting microplastics ingestion in the freshwater mussel Dreissena polymorpha. Using polystyrene spheres (5, 10, 45, 90 µm), we determined the body burden of microplastics in the mussels in relation to (1) exposure and depuration time, (2) body size, (3) food abundance and (4) MP concentrations. D. polymorpha rapidly ingested microplastics and excreted most particles within 12 h. Few microplastics were retained for up to one week. Smaller individuals had a higher relative body burden of microplastics than larger individuals. The uptake of microplastics was concentration‐dependent, while an additional food supply (algae) reduced it. We also compared the ingestion of microplastics by D. polymorpha with two other freshwater species (Anodonta anatina, Sinanodonta woodiana) highlighting that absolute and relative uptake depends on the species and the size of the mussels. In addition, we determined toxicity of polystyrene fragments (≤ 63 µm, 6.4–100,000 p mL‐1) and diatomite (natural particle, 100,000 p mL‐1) in D. polymorpha after 1, 3, 7 and 42 days of exposure investigating the clearance rate, energy reserves and oxidative stress. Despite ingesting large quantities, exposure to polystyrene fragments only affected the clearance rate of D. polymorpha. Further, results of the microplastic and diatomite exposure did not differ significantly. D. polymorpha, therefore, is unaffected by or can compensate for polystyrene fragment toxicity even at concentrations beyond current environmental levels. This article is protected by copyright. All rights reserved.

Full text

wileyonlinelibrary.com/ETC © 2021 The Authors
Environmental Toxicology and Chemistry—Volume 40, Number 8—pp. 2247–2260, 2021
Received: 9 February 2021
|
Revised: 16 March 2021
|
Accepted: 17 April 2021 2247
Environmental Toxicology
Ingestion and Toxicity of Polystyrene Microplastics in
Freshwater Bivalves
Annkatrin Weber,
a
Nina Jeckel,
a
Carolin Weil,
a
Simon Umbach,
a
Nicole Brennholt,
b
Georg Reifferscheid,
b
and Martin Wagner
c,
*
a
Faculty of Biological Sciences, Department of Aquatic Ecotoxicology, Goethe University Frankfurt am Main, Frankfurt am Main, Germany
b
Department of Biochemistry and Ecotoxicology, Federal Institute of Hydrology, Koblenz, Germany
c
Department of Biology, Norwegian University of Science and Technology, Trondheim, Norway
Abstract: The ubiquity of microplastics in aquatic ecosystems has raised concerns over their interaction with biota. However,
microplastics research on freshwater species, especially mollusks, is still scarce. We, therefore, investigated the factors
affecting microplastics ingestion in the freshwater mussel Dreissena polymorpha. Using polystyrene spheres (5, 10, 45,
90 µm), we determined the body burden of microplastics in the mussels in relation to 1) exposure and depuration time, 2)
body size, 3) food abundance, and 4) microplastic concentrations. D. polymorpha rapidly ingested microplastics and ex-
creted most particles within 12 h. A few microplastics were retained for up to 1 wk. Smaller individuals had a higher relative
body burden of microplastics than larger individuals. The uptake of microplastics was concentration‐dependent, whereas an
additional food supply (algae) reduced it. We also compared the ingestion of microplastics by D. polymorpha with 2 other
freshwater species (Anodonta anatina,Sinanodonta woodiana), highlighting that absolute and relative uptake depends on
the species and the size of the mussels. In addition, we determined toxicity of polystyrene fragments (≤63 µm,
6.4–100 000 p mL
–1
) and diatomite (natural particle, 100 000 p mL
–1
)inD. polymorpha after 1, 3, 7, and 42 d of exposure,
investigating clearance rate, energy reserves, and oxidative stress. Despite ingesting large quantities, exposure to poly-
styrene fragments only affected the clearance rate of D. polymorpha. Further, results of the microplastic and diatomite
exposure did not differ significantly. Therefore, D. polymorpha is unaffected by or can compensate for polystyrene fragment
toxicity even at concentrations above current environmental levels. Environ Toxicol Chem 2021;40:2247–2260. © 2021 The
Authors. Environmental Toxicology and Chemistry published by Wiley Periodicals LLC on behalf of SETAC.
Keywords: Microplastics; Toxic effects; Mollusk toxicity
INTRODUCTION
Plastics are part of almost every aspect of modern human
life. However, the rising global plastic production
(PlasticsEurope 2018) coincides with plastic pollution in nature.
In particular, the fragmentation of plastic debris results in a
global distribution of micrometer‐sized plastic particles, so‐
called microplastics (1–1000 µm; Hartmann et al. 2019), in the
aquatic environment. Accordingly, there is emerging concern
over the potential environmental impacts of microplastics.
In recent years, research on microplastic exposure and
toxicity has especially focused on marine bivalves. Their high
filtration activity results in a higher ingestion of microplastics
compared to other taxa (Setälä et al. 2016), rendering bivalves
especially susceptible to microplastic exposure. Uptake of mi-
croplastics has repeatedly been demonstrated in wild and
cultured marine bivalves (Li J et al. 2019). The current micro-
plastic body burden varies intensively from ≤1 (Railo et al.
2018) up to several hundred particles per individual (Mathalon
and Hill 2014). Numerous experimental studies further confirm
that bivalves ingest microplastics of different size, shape, and
polymer type (see Brillant and MacDonald 2000; Bråte et al.
2018; Li L et al. 2019).
While there is sufficient evidence demonstrating that bi-
valves ingest microplastics, less is known regarding the in-
gestion kinetics. In theory, microplastic ingestion will depend
on multiple factors, including the physicochemical properties
of microplastics (e.g., size, shape, polymer type), the exposure
and depuration time of the individuals, biological traits of
This article includes online‐only Supplemental Data.
This is an open access article under the terms of the Creative Commons
Attribution‐NonCommercial License, which permits use, distribution and
reproduction in any medium, provided the original work is properly cited and is
not used for commercial purposes.
Published online 30 April 2021 in Wiley Online Library
(wileyonlinelibrary.com).
DOI: 10.1002/etc.5076
* Address correspondence to martin.wagner@ntnu.no

species (e.g., feeding type), microplastic bioavailability, and
the presence of other particulate matter. Previous pub-
lications investigated some of these relevant factors in mus-
sels (Brillant and MacDonald 2000; Capolupo et al. 2018;
Woods et al. 2018; Fernández and Albentosa 2019; Gonçalves
et al. 2019; Rist et al. 2019a, 2019b). However, the use of
different experimental designs (e.g., different types of mi-
croplastics and exposure periods) limits the comparability of
the existing data and prevents a general assessment of mi-
croplastic uptake kinetics. Furthermore, understanding better
the mussel microplastic ingestion kinetics will improve our
understanding of a potential microplastic uptake by humans
through mussels as a food source (Mercogliano et al. 2020).
Yet, understanding the factors driving microplastic ingestion
by bivalves is important to identify species that are particularly
susceptible to microplastic exposure and to design adequate
toxicity experiments.
To address this knowledge gap, we studied the ingestion
and egestion of microplastics in the freshwater species
Dreissena polymorpha in 4 experiments with similar design. For
this, we exposed D. polymorpha to a mixture of 5‐,10‐,45‐,
and 90‐µm polystyrene (PS) spheres to analyze how the factors
exposure and depuration time (experiment 1), body size (ex-
periment 2), food abundance (experiment 3), and microplastic
concentration (experiment 4) affect the body burden of mi-
croplastics in D. polymorpha.
Dreissenids are a well‐established test organism in fresh-
water ecotoxicology (Péden et al. 2019), but because of their
small size and invasive character, they may not be repre-
sentative of larger species (e.g., Unionidae in European fresh-
water systems). We, therefore, repeated ingestion experiments
1 and 2 with the 2 larger freshwater species Anodonta anatina
and Sinanodonta woodiana to evaluate species‐and size‐
specific variations.
In addition to microplastic ingestion, further data are
needed regarding the toxicity of microplastics in freshwater
bivalves. Previous research on Corbicula fluminea (Rochman
et al. 2017; Guilhermino et al. 2018; Oliveira et al. 2018;
Baudrimont et al. 2020) and D. polymorpha (Magni et al. 2018,
2019, 2020; Binelli et al. 2020) provided contradicting results.
Although some of these studies identified significant
microplastic‐induced neurotoxicity, oxidative stress, and a
change in feeding behavior, others did not observe such ef-
fects. Thus, we exposed D. polymorpha over 1, 3, 7, and 42 d
to 6.4 to 100 000 particles (p) mL
–1
using PS fragments
(≤63 µm) and analyzed effects on the clearance rate as well as
energy reserves (protein, glycogen, lipid content) and oxidative
stress levels in the midgut gland (synonyms “digestive gland,”
“hepatopancreas”)ofD. polymorpha. In addition, we exposed
the mussels to diatomite (100 000 p mL
–1
) to compare the
toxicity of microplastics and naturally occurring particles.
MATERIALS AND METHODS
Mussel culture
Dreissena polymorpha were collected from Oberwald Lake in
Mörfelden‐Walldorf, Germany (49°59′0.242″N, 8°35′48.666″E).
Sinanodonta woodiana and A. anatina were purchased from
local fish shops and cultures. In the laboratory, all bivalve species
were cultured in aerated Organisation for Economic Co‐
operation and Development (OECD) medium (Organisation for
Economic Co‐operation and Development 2016) at 14 °C water
temperature and a 16:8‐h light:dark cycle. Sinanodonta wood-
iana and A. anatina were kept in 150‐L tanks with up to
30 individuals tank
–1
.Dreissena polymorpha was cultured in a
50‐L tank with approximately 200 individuals tank
–1
.Mussels
used in the ingestion experiments were allowed to acclimatize
for at least 1wk, whereas those used in the toxicity study
(D. polymorpha) were cultured for at least 4 wk prior to the ex-
periments. Twice a week, at least half of the medium was re-
newed. Mussels were fed with algae (D. subspicatus)adlibitum
at least thrice a week.
Particle characterization
For the ingestion experiments, we used 5‐,10‐,45‐,and
90‐µmplain,fluorescent PS spheres. We used spheres with a
homogenous size because this allows for investigation of the size
dependency of microplastic ingestion and depuration. The 10‐,
45‐, and 90‐µm PS spheres were purchased from PolyScience
(Fluoresbrite YG microspheres; excitation 441 nm, emission
486 nm). The 5‐µm spheres were obtained from MagSphere
(excitation 538 nm, emission 584 nm). We suspended the PS
spheres in ultrapure water (microplastic stock suspensions) and
determined the particle concentration and size distribution in
the stock suspensions with a Coulter counter (Multisizer 3;
Beckman Coulter; details in Supplemental Data, S1).
For the toxicity study, we used PS fragments (≤63 µm) be-
cause of the higher environmental relevance of fragments
compared to spheres (De Sá et al. 2018) as well as diatomite
(Sigma‐Aldrich). The PS fragments were prepared from orange
fluorescent drinking cups (excitation 360–370 nm) by cryomil-
ling, followed by sieving with a 63‐µm sieve. The diatomite
particles were sieved in the same way to obtain the size fraction
≤63 µm. In our previous work, we confirmed that the cups were
made of PS and had low but detectable concentrations of
chemicals which could not be matched to substances com-
monly used in plastics (see Weber et al. [2020] for details).
Further, scanning electron microscopic images of PS fragments
as well as diatomite particles are published in Weber et al.
(2021). These images indicate that both microplastic and dia-
tomite powder included nano‐sized particles. As stated in
Weber et al. (2021), nanoparticle abundance was not quantified
because micro‐sized particles would have blocked the nano-
particle tracking analysis instrument.
We determined the particle number per powder mass
(number mg
–1
) and the particle size distribution with a Coulter
counter (size range 2–60 µm). The microplastic and the dia-
tomite powder contained 287 526 and 4 632 990 p mg
–1
, re-
spectively. Particle size distribution of the PS and the diatomite
suspensions increased exponentially with decreasing particle
size; 90% of the microplastic and the diatomite particles were
smaller than 12.4 and 11.8 µm, respectively (for detailed
methods and results see Weber et al. [2020]).
2248 Environmental Toxicology and Chemistry, 2021;40:2247–2260—A. Weber et al.
© 2021 The Authors wileyonlinelibrary.com/ETC

Relevant factors affecting microplastic ingestion
and depuration by D. polymorpha
Basic exposure scenario. All ingestion experiments with
D. polymorpha were performed using the same basic exposure
scenario (see Supplemental Data, S2.1). Ten‐liter glass tanks
(19 ×29.3 ×19 cm) were filled with 7.5 L aerated OECD me-
dium and quartz sand (5 cm layer, previously annealed at
200 °C for 24 h). Each tank was equipped with a pump (Tetra
IN400plus; Tetra) set to the lowest rate to create a constant
circular water flow and to keep the particles suspended in the
water column. The extent to which the particles remained in the
water phase was characterized in an extra experiment. Despite
a constant water flow, PS spheres settled, with smaller micro-
plastics remaining longer in the water column. From 6 h on-
ward, only 5‐and 10‐µm PS spheres were present in the water
phase; and after 12 h, 50.2% of the 5‐µm spheres and 77.3% of
the 10‐µm spheres had cleared from the water phase (for de-
tails on methods and results, see Supplemental Data, S2.2).
Twelve hours prior to the start of the experiments, 6 mussels
were transferred to each tank. Algae (D. subspicatus) were
added 1 h before the start of the experiment to stimulate fil-
tration behavior. Each experiment started when microplastics
were added to the tanks by pipetting the microplastic stock
suspensions directly below the water surface into the flow of
the pump. Because of the short exposure times (1–48 h), the
water was not exchanged during the ingestion experiments.
Throughout each experiment, we visually monitored each
mussel hourly and recorded whether its valves were opened or
closed. Only mussels which were open at least at 50% of the
monitored time points (e.g., ≥6 time points throughout a 12‐h
exposure period) were classified as “active mussels“and fur-
ther analyzed. In the following sections, we used this basic
exposure scenario to examine the impact of the factors
exposure and depuration time, body size, food abundance,
and exposure concentration on microplastic ingestion by
D. polymorpha.
Exposure and depuration time (experiment 1). The im-
pact of the exposure time on microplastic ingestion was ana-
lyzed by exposing D. polymorpha (1.8–2.2 cm maximal shell
length) for 1, 3, 6, 12, 24, and 48 h to a mixture of 5‐,10‐, and
45‐µm PS spheres at 3 p mL
–1
each. Further, we added 90‐µm
PS spheres (0.1 p mL
–1
) to examine whether D. polymorpha is
able to ingest also larger microplastic particles. We applied
these at lower concentrations because of high material costs,
and the respective data were, thus, analyzed separately. We
used one separate tank with 6 individuals for each time point.
Algae were added to each tank at a concentration of 1 mg L
–1
total organic carbon (TOC). Four “active”individuals from each
tank were analyzed for their microplastic body burden.
For analysis of the impact of depuration time on the body
burden of microplastics in D. polymorpha, we exposed
30 mussels (5 tanks with 6 individuals each) in the presence of
algae (1 mg L
–1
) to PS spheres (sizes and concentrations as
stated in the previous paragraph) for 12 h. After the exposure,
28 active mussels were randomly selected and transferred into
tanks filled with microplastic‐free OECD medium. Individual D.
polymorpha were held there for 1, 3, 6, 12, 24, 72, and 168 h
(one tank per depuration time point with 4 individuals each).
Directly after the transfer and subsequently every 24 h after-
ward, mussels were fed 1 mg L
–1
TOC algae. Because mussels
are able to reingest excreted microplastic, we transferred the
mussels into new tanks with fresh medium after 24, 72, and
120 h to minimize microplastic reuptake.
Body size (experiment 2). In the second experiment, we
evaluated the relationship between body size and microplastics
in D. polymorpha. For this, D. polymorpha individuals from
3 different size classes (1.0–1.5, 1.8–2.2, and 2.5–3.0 cm) were
exposed in the presence of algae (1 mg L
–1
TOC) to PS spheres
(sizes and concentrations as above, see section Exposure and
depuration time [experiment 1]) for 12 h. For each size class, 2
tanks were set up with 6 individuals each. After the exposure,
microplastic body burden was analyzed in 8 active out of the
12 exposed individuals per size class.
Food abundance (experiment 3). We evaluated how algae
abundance affects the microplastic body burden by exposing
D. polymorpha (1.8–2.2 cm) for 12 h to PS spheres (sizes and
concentrations as above, see section Exposure and depuration
time [experiment 1]) in the presence of 3 algae concentrations
(0.2, 1, or 5 mg L
–1
TOC algae). The microplastic to algae ratios
(based on particle numbers) in the 3 exposures were 1:5589,
1:27 798, and 1:135 477, respectively, based on the sum con-
centration of 5‐,10‐,45‐, and 90‐µm spheres and of algae cells
in each tank. Algae were added 1 h before the start of the
experiment to stimulate filtration activity. For each algae con-
centration, 2 tanks with 6 mussels each were established and 8
active individuals per treatment analyzed.
Microplastic concentration (experiment 4). We inves-
tigated the relationship of microplastic concentration and mi-
croplastic body burden in D. polymorpha (2.5–3.0 cm) by
exposing the mussels in the presence of algae (1 mg L
–1
TOC)
for 12 h to 5‐,10‐, and 45‐µm PS spheres at either 0.3 or
3pmL
–1
each. Again, we also added 90‐µm PS spheres (0.01 or
0.1 p mL
–1
, respectively). The 10‐fold lower concentration of
0.3 p mL
–1
was chosen to resemble environmental concen-
trations already reported for freshwater systems (Leslie et al.
2017; Lahens et al. 2018). For both concentrations, 2 exposure
tanks were prepared (6 mussels per tank) and 8 active mussels
were analyzed per concentration.
Analysis of microplastic body burden in D. polymorpha.
After each experiment, D. polymorpha individuals were thor-
oughly rinsed with tap water and frozen at –80 °C. After de-
frosting, the shells were removed. All tissues were lyophilized to
determine the total dry weight of each mussel. Afterward, tis-
sues were lysed in 20 to 40 mL 10% potassium hydroxide sol-
ution at 55°C for 24 to 48 h. The lysate was filtered on glass fiber
filters (pore size 1.25 µm; VWR). Each filter was analyzed visually
with a fluorescence microscope (BX50, ×40 magnification;
Polystyrene microplastics in freshwater bivalves—Environmental Toxicology and Chemistry, 2021;40:2247–2260 2249
wileyonlinelibrary.com/ETC © 2021 The Authors

Olympus), and the number of fluorescent spheres on the whole
filter was determined for each microplastic type (for details see
Supplemental Data, S2.2). The body burden in the mussels was
characterized both separately for each microplastic type and as
the total number of ingested microplastics (total body burden).
The latter corresponds to the sum of 5‐,10‐,and45‐µmPS
spheres. Because of the divergent exposure concentrations (see
section Exposure and depuration time [experiment 1]), the re-
sults for the 90‐µm spheres are not included in the total micro-
plastic body burden and are presented separately.
Quality assurance. The background contamination with
fluorescent particles in D. polymorpha tissues was determined
by analyzing mussels (1.0–1.5, 1.8–2.2, 2.5–3.0 cm) from the
culture which had not been exposed to microplastics. We lysed
3D. polymorpha individuals from each size class, as described
in Analysis of microplastics body burden in D. polymorpha, and
corrected all data from the ingestion experiments for the mi-
croplastic body burden in those control mussels (see Supple-
mental Data, S3.2) by subtracting the average number of
microplastic‐resembling particles per control mussel from the
microplastic body burden in exposed individuals (separate data
correction for each particle type).
Comparison of microplastic ingestion between
freshwater mussel species
We further compared microplastic ingestion by
D. polymorpha with other freshwater mussel species. Originally,
we intended to repeat the experiments just described with the
native species A. anatina. However, the number of available
A. anatina specimens was too low. Therefore, we used a second
species (S. woodiana) with similar morphology and ecology and
limited the comparative studies to experiments 1 and 2.
We repeated experiment 1 with S. woodiana (9.5–12.0 cm).
For the depuration experiment, we exposed 24 mussels
(4 tanks with 6 individuals each). After the exposure, 16 “ac-
tive”S. woodiana (for definition, see section Basic exposure
scenario) were randomly selected and transferred into tanks
filled with microplastic‐free OECD medium. We limited the
number of depuration time points to 4 (12, 24, 72, and 168 h)
because of the limited number of mussels available.
We repeated experiment 2 with S. woodiana and A. anatina
(2 size classes each: 6.0–8.0, 9.5–12.0 cm). For each species
and size class, 2 tanks were set up with 6 individuals each. After
the exposure, microplastic body burden was analyzed in 8 ac-
tive out of the 12 exposed individuals from each treatment.
Because of the mussels' inactivity in some treatments, the ex-
periment was repeated with a third tank to obtain 8 active
individuals per treatment.
For analysis of the microplastic body burden in S. woodiana
and A. anatina, we did not analyze the whole body as we did
for D. polymorpha because of the large body size and, thus,
insufficient tissue lysis. Instead, we removed the mantle, gills,
and foot. In a prior experiment, we demonstrated that micro-
plastic levels in the mantle, gills, and foot were low compared
to the other tissues; and we, thus, consider the number of
microplastics in the removed tissues negligible (for details see
Supplemental Data, S3.1). After dissection, we lyophilized both
the mantle, gills, and foot and the remaining tissues to de-
termine the total dry weight of each mussel. The mantle, gills,
and foot tissue was discarded afterward, whereas the re-
maining tissue was analyzed for the microplastic body burden
as described in Analysis of microplastics body burden in D.
polymorpha. Again, we determined background contamination
by lysing control mussels and correcting the results from the
ingestion experiments accordingly (for detailed results, see
Supplemental Data, S3.2).
Microplastic toxicity in D. polymorpha
Exposure scenario. Dreissena polymorpha were exposed to
either 6.4, 160, 4000, or 100 000 p mL
–1
PS fragments (≤63 µm)
or 100 000 p mL
–1
diatomite for 1, 3, 7 (acute exposure), and
42 d (chronic exposure). We also included a negative control
without microplastic/diatomite. Each particle concentration was
tested in a separate glass tank (14 ×20 ×20 cm) with 3 L of
OECD medium and 40 mussels (2.0–2.3 cm). To keep the ex-
periments manageable, we set up one set of 6 glass tanks (one
tank per treatment) for the acute exposures (1, 3, and 7 d; for all
3 time points mussels were sampled from the same tank),
whereas for the chronic exposure we set up a separate set of
6 glass tanks (42 d; scheme of the experimental design in
Supplemental Data, Figure S2).
The required masses of microplastic and diatomite were
weighed for each treatment and added directly to the medium
in the aquaria. For the 6.4 p mL
–1
treatment, we used a 100‐fold
diluted stock suspension in OECD medium, which was applied
to the respective tank. Each tank was constantly aerated
through 2 glass pipettes to enhance particle dispersion in the
water phase. The mussels were fed with algae (D. subspicatus,
0.25 mg TOC individual
–1
) daily. In the chronic exposure ex-
periment, the medium was completely renewed every 7 d by
transferring the mussels to new tanks prepared as described.
Mortality was recorded daily, and dead individuals were
removed. After 1, 3, 7, and 42 d, the clearance rate of 10 in-
dividuals per treatment was determined. In case of acute
exposures, individuals were reintroduced into their corre-
sponding tanks afterward. Further, at each time point,
10 individuals were frozen in liquid nitrogen and stored at
–80 °C for energy reserve and stress metabolite analysis.
Clearance rate, energy reserves, and stress metabo-
lites. We quantified the clearance rate of D. polymorpha by
placing 10 mussels per treatment and time point individually in
an algae suspension and determining algae concentrations
in the medium as chlorophyll fluorescence (in relative
fluorescence units [RFUs]) prior to and after 45 min (Tecan;
GENios; excitation 440 nm, emission 680 nm). The starting
concentration was 4000 ±360 RFU (±standard deviation), rep-
resenting 1.21 ×10
7
±1.08 ×10
6
algae cells mL
–1
. The clear-
ance rate is the difference in RFUs before and after 45 min. We
used Raphidocelis subcapitata (formerly Pseudokirchneriella
2250 Environmental Toxicology and Chemistry, 2021;40:2247–2260—A. Weber et al.
© 2021 The Authors wileyonlinelibrary.com/ETC

subcapitata) instead of D. subspicatus because it is unicellular,
allowing for more accurate fluorescence analyses.
As biochemical endpoints, we analyzed the energy content
as well as oxidative stress markers in the midgut gland of
D. polymorpha. The midgut glands from 10 individuals per
treatment were dissected, wet‐weighed, homogenized, and
frozen at –80 °C. Energy reserves in the midgut gland homo-
genates were measured as the protein content according to
Bradford (1976) and as the glycogen content (anthrone assay)
and the total lipid content (sulfo‐phospho‐vanillin assay) ac-
cording to Benedict (2014). Oxidative stress was quantified as
the malondialdehyde (MDA) content (an important biomarker
for lipid peroxidation) as well as the remaining antioxidant ca-
pacity. Concentrations of MDA were measured using the thi-
obarbituric acid reactive substances (TBARS) assay (Hodges
et al. 1999; Furuhagen et al. 2014). The remaining antioxidant
capacity in the midgut gland was measured with the oxygen
radical absorbance capacity (ORAC) assay (Ou et al. 2001;
Furuhagen et al. 2014). Further methodological details are
provided in Weber et al. (2020). Because of highly divergent
results for midgut gland with a wet weight <5 mg compared to
midgut gland >5 mg, we excluded results on energy reserves
and oxidative stress from individuals with an midgut gland
<5mg (n=7–10).
Statistics
All statistical analyses were performed with IBM SPSS Sta-
tistics (Ver 25) using one‐way or 2‐way analysis of variances
(ANOVAs). Prior to each analysis, we tested for normality
(Shapiro‐Wilks test), variance homogeneity (Levene test), and
heteroscedasticity (Ftest). Very few treatments violated the
normality criteria. In this case, we reperformed the statistical
analysis after outlier exclusion but did not observe changes in
the results. All data were visualized using GraphPad Prism 8.4.3
(GraphPad Software).
Statistics for the ingestion experiments with D. poly-
morpha. We analyzed the effects of the factors exposure time
and depuration time (experiment 1), individual size (experiment
2), food abundance (experiment 3), and microplastic concen-
tration (experiment 4) on the microplastic body burden
by applying separate one‐way ANOVAs for each factor
(dependent variable, absolute or relative microplastic body
burden; fixed variables, factors listed). In addition, we per-
formed Tukey's posttests to analyze differences between the
treatments of each experiment (full‐factorial comparison of all
treatments). Data for total microplastic body burden were log‐
transformed (depuration time, individual size, microplastic
concentration), square root–transformed (food abundance), or
not transformed (exposure time) for the statistical analysis.
Statistics for the comparison of microplastic ingestion
between species. We analyzed the effects of the variable
speciesincombinationwiththevariableexposuretime,
depuration time (experiment 1), or individual size (experiment
2) as well as their interaction on the total microplastic body
burden (absolute or relative) in the analyzed freshwater mus-
sels with 2‐way ANOVAs. Both variables were integrated as
fixed variables, whereas the total microplastic body burden
was used as a dependent variable. Further, for the results of
experiment 2 we performed Tukey's posttest to determine
homogenous subgroups in regard to the different species
(D. polymorpha,S. woodiana,A. anatina)aswellassize
classes (1.0–1.5, 1.8–2.2, 2.5–3.0, 6.0–8.0, 9.5–12.0 cm). Data
for total microplastic body burden were log‐transformed
(depuration time, individual size) or square root–transformed
(exposure time) for the statistical analysis.
Statistics for the microplastic toxicity study with
D. polymorpha. In the toxicity study, the effects of micro-
plastic concentration and exposure time (both fixed variables)
as well as their interaction (microplastic concentration ×
exposure time) were determined with 2‐way ANOVAs for each
endpoint (clearance rate, protein, glycogen, total lipids, MDA,
antioxidative capacity) as a dependent variable. Data for the
dependent variable were integrated as either log‐(glycogen,
lipids, MDA [TBARS], Trolox equivalents [ORAC]), square root–
(protein), or third root–transformed (RFU [clearance rate]).
Statistical comparison of the microplastic and the diatomite
exposure (both 100 000 p mL
–1
)with2‐way ANOVAs was
performed as described for the toxicity study but with
thevariableparticletypeinsteadofmicroplasticconcen-
tration. Data for the dependent variable were transformed as
described in the previous paragraph.
RESULTS
Factors affecting microplastic ingestion and
depuration by D. polymorpha
Exposure and depuration time (experiment 1). Exposure
time significantly affects total microplastic number in D.
polymorpha (p<0.01). The total microplastic body burden
(5‐,10‐,45‐µm PS spheres) in D. polymorpha was highest after
1 h and decreased afterward, with another peak after 12 h
(Figure 1A). These 2 peaks were significantly higher compared
to 48 h of exposure (p<0.05). The 10‐µm PS spheres were
found in highest quantities except after 12‐h exposure, when
the quantity of 5‐µm PS spheres exceeded that for 10‐µm
spheres.
Similarly, the depuration time had a significant effect on
microplastic number (p<0.001). The total microplastic body
burden significantly decreased from 1 h of depuration onward
compared to the 12‐h exposure without any depuration
phase (12(+0) h). Further, the total body burden after 1 h of
depuration was still significantly higher compared to 3, 24, 72,
and 168 h of depuration (p<0.05). The decrease was most
distinct for 5‐and 10‐µm PS spheres (Figure 1B). After 7 d,
microplastic numbers had decreased to 0.3% (0.5 p
individuals
–1
[median]) of the original body burden after
12(+0) h (151.0 p individuals
–1
). We did not detect 90‐µm
spheres in D. polymorpha tissues in experiment 1 or in any of
the following experiments.
Polystyrene microplastics in freshwater bivalves—Environmental Toxicology and Chemistry, 2021;40:2247–2260 2251
wileyonlinelibrary.com/ETC © 2021 The Authors

FIGURE 1: Impact of exposure and depuration time (A,B) as well as body size (C,D) on microplastic (MP) ingestion by Dreissena polymorpha.
Mussels were (A) exposed to microplastic (5‐,10‐, and 45‐µm polystyrene [PS] spheres, 3 p mL
–1
each; 90‐µm PS spheres, 0.1 p mL
–1
) for 1 to 48 h or
(B) exposed for 12 h (12 (+0)) and then transferred to microplastic‐free medium for up to 168 h. n=4 for each time point. Data points indicate
median values. (C,D) Mussels from 3 size classes were exposed to microplastic for 12 h, and the (C) absolute total body burden (sum of 5‐,10‐, and
45‐µm spheres) as well as the (D) relative total body burden (per dry wt) were determined (n=8). Lines indicate the median. Statistics: one‐way
analysis of variance with Tukey's posttest; different letters indicate significant differences between the treatments. No 90‐µm spheres were detected
in D. polymorpha in any of the exposure experiments.
FIGURE 2: Impact of food abundance (A) and microplastic (MP) concentration (B) on microplastic ingestion by Dreissena polymorpha.(A) Mussels
were exposed to microplastic (5‐,10‐, and 45‐µm polystyrene [PS] spheres, 3 p mL
–1
each; 90‐µm PS spheres, 0.1 p mL
–1
) for 12 h in the presence of
3 algae concentrations (n=8, data points indicate median). (B) Mussels were exposed to microplastic (5‐,10‐,45‐, and 90‐µm PS spheres) for 12 h at
concentrations of 0.3 and 3 p mL
–1
each (90‐µm spheres, 0.01 and 0.1 p mL
–1
) for 12 h. Data points represent the total body burden (sum of 5‐,10‐,
and 45‐µm PS spheres) per individual. The line indicates the median. Statistics: one‐way analysis of variance with Tukey's posttest; different letters
indicate significant differences between the treatments. No 90‐µm spheres were detected in D. polymorpha in any of the exposure experiments.
TOC =total organic carbon.
2252 Environmental Toxicology and Chemistry, 2021;40:2247–2260—A. Weber et al.
© 2021 The Authors wileyonlinelibrary.com/ETC

Body size (experiment 2). Ingestion of microplastics by D.
polymorpha varied considerably within the 3 tested size
classes. The absolute microplastic body burden in individuals
of the 3 size classes did not differ significantly (Figure 1C;
p>0.05). When considering the relative body burden (micro-
plastic number per dry wt), the smallest individuals had sig-
nificantly higher microplastic numbers compared to larger
individuals (Figure 1D; p<0.01).
Food abundance (experiment 3). Dreissena polymorpha
ingested a lower amount of microplastic when more food was
available. The decrease of the total body burden was, however,
not linear but most pronounced between 0.2 and 1 mg TOC L
–1
algae (equivalent to 5.08 ×10
4
and 2.53 ×10
5
algae cells L
–1
;
Figure 2A). The mussels did not take up 45‐µm spheres when
fed 5 mg TOC L
–1
algae.
Microplastic concentration (experiment 4). In D. poly-
morpha, the total body burden increased when exposing the
mussels to a 10‐fold higher microplastic concentration, the in-
crease was, however, not significant (p>0.05; Figure 2B).
Further, the increase was not proportional: a 10‐fold increase in
microplastic concentrations resulted only in a 5.1‐fold (30.7 vs
156.9 microplastic individual
–1
) higher microplastic burden.
Comparison of microplastic ingestion between
freshwater mussel species
We reperformed experiment 1 (factors exposure and dep-
uration time) with S. woodiana as a second species and com-
pared these data with the results for D. polymorpha to
determine species‐specific differences (Figure 3A and B).
Considering the microplastic ingestion after various exposure
periods, total body burden (sum of 5‐,10‐,45‐µm PS spheres)
differed significantly between S. woodiana and D. polymorpha
(p<0.001), with S. woodiana ingesting much more microplastic
than D. polymorpha (Figure 3A). In S. woodiana, the body
burden peaked after 3 and 6 h, whereas in D. polymorpha
microplastic levels peaked after 1 and 12 h. Consequently, no
significant effect of exposure time was observed (p>0.05). The
same applies to the interaction of both factors (p>0.05).
In the depuration experiments, both the variables species
(p<0.001) and depuration time (p<0.001) as well as their in-
teraction (p<0.01) significantly affected microplastic body
burden in the mussels. In both species, microplastic numbers
reduced with increasing depuration time (Figure 3B). However,
microplastic depuration was faster in D. polymorpha with al-
most complete microplastic clearance within 12 h, whereas the
lowest body burden in S. woodiana was reached after 72 h.
After 7 d of depuration, microplastic clearance in both species
was similar (D. polymorpha, 99.7%; S. woodiana, 95.8%).
However, whereas in D. polymorpha only few microplastic
particles remained in the mussel tissues, we detected >100
microplastics in S. woodiana.
When comparing absolute body burdens in D. polymorpha,
A. anatina, and S. woodiana of various sizes (experiment 2;
Figure 3C), neither body size nor species nor their interaction
had a significant effect. The Tukey's posttest, however, showed
that D. polymorpha as well as the 3 smaller size classes
(1.0–1.5, 1.8–2.2, and 2.5–3.0 cm) form homogenous sub-
groups which differ from the remaining subgroups consisting of
A. anatina and S. woodiana as well as the 2 larger size classes
(6.0–8.0 and 9.5–12.0 cm). Absolute microplastic ingestion by
D. polymorpha was, therefore, lower compared to the 2 larger
freshwater species.
With regard to relative body burden in the freshwater
mussels (Figure 3D), the variables individual size (p<0.001)
and species (p<0.05) had a significant effect, although no in-
teraction was observed (p>0.05). For the variable species, all
3 species formed separate subgroups (Tukey's posttest), with
D. polymorpha having the highest and A. anatina the lowest
microplastic number in its tissues. With regard to individual
size, 3 subgroups were identified (1.0–1.5 cm; 1.8–2.2, 2.5–3.0,
and 6.0–8.0 cm; and 6.0–8.0 and 9.5–12.0 cm), which indicates
that the relative body burden increases with decreasing in-
dividual size.
All 3 species ingested high proportions of 10‐µmspheres
(Figure 3E). Also, D. polymorpha (especially 1.0–1.5 cm) ingested
large quantities of 5‐µm spheres, whereas in the larger species,
S. woodiana and A. anatina,45‐µm spheres were more abundant.
Thus, smaller mussel species seem to ingest smaller micro-
plastics. The same trend applied to differently sized individuals
within the 3 species (except for D. polymorpha,1.8–2.2 cm). The
burden of smaller microplastics (5 and 10 µm) increased with
smaller body size. In contrast to D. polymorpha,the2larger
species also ingested 90‐µm PS spheres. After a 12‐hexposure,
we detected up to 37 (6.0–8.0 cm) and 49 (9.5–12.0 cm) 90‐µmPS
spheres in A. anatina as well as up to 47 (6.0–8.0 cm) and
56 (9.5–12.0 cm) 90‐µm spheres in S. woodiana.
Microplastic toxicity in D. polymorpha
No mortality occurred in any treatment during the acute
exposure (1, 3, and 7 d) of D. polymorpha to PS fragments or
diatomite. In the chronic exposure (42 d), mortality remained
low, with 0% in the 100 000 p mL
–1
microplastic treatment, with
10% in the control, 160 and 4000 p mL
–1
microplastic treat-
ment, as well as with 12.5% in the 6.4 p mL
–1
microplastic and
100 000 p mL
–1
diatomite treatment.
The exposure of D. polymorpha to microplastics
(6.4–100 000 p mL
–1
) for 1, 3, 7, and 42 d did not induce sig-
nificant effects (p>0.05) on the energy reserves (proteins,
glycogen, lipids) or oxidative stress levels (MDA content, re-
maining antioxidant capacity) in the midgut gland of the ana-
lyzed individuals (Figure 4 and Table 1). However, we observed
significant microplastic effects on the clearance rate of D.
polymorpha (p<0.05; Table 1), with a pronounced increase in
the 100 000 p mL
–1
treatment after 7 and 42 d of microplastic
exposure (Figure 4F). In contrast, the exposure time sig-
nificantly affected the energy reserves and oxidative stress level
in D. polymorpha (p<0.01; Table 1). However, there was no
clear linear trend between the exposure time and the end-
points except for the remaining antioxidative capacity, which
decreased with increasing exposure time (Figure 4E).
Polystyrene microplastics in freshwater bivalves—Environmental Toxicology and Chemistry, 2021;40:2247–2260 2253
wileyonlinelibrary.com/ETC © 2021 The Authors

In addition, we statistically compared the 100 000 p mL
–1
microplastic and diatomite exposures for each endpoint con-
sidering the variables particle type and exposure time as well as
their interaction. Significant differences were only observed for
the exposure time (p<0.05; Supplemental Data, Table S4)
regarding the protein content, MDA (TBARS), and the anti-
oxidative capacity (ORAC). The particle type did not cause
significant effects (p>0.05). Particle effects due to an micro-
plastic and diatomite exposure did not, therefore, differ in
D. polymorpha.
FIGURE 3: Comparison of the total body burden of microplastics (MP) in Sinanodonta woodiana and Dreissena polymorpha after different
exposure (A) or depuration (B) time periods as well as comparison of the (C) absolute (per individual) and (D) relative body burden (total microplastic
number per dry wt) in D. polymorpha,S. woodiana, and Anodonta anatina of various body sizes (12‐h exposure). (E) Ratios of the microplastic types
in mussels of different size classes (12‐h exposure). (A,B)Sinanodonta woodiana and D. polymorpha were exposed as described in Figure 1A and B
(n=4, data points indicate median). (C–E)Sinanodonta woodiana,A. anatina, and D. polymorpha were exposed as described in Figure 1C and D
(n=8, black circles indicate geometric mean).
2254 Environmental Toxicology and Chemistry, 2021;40:2247–2260—A. Weber et al.
© 2021 The Authors wileyonlinelibrary.com/ETC

DISCUSSION
Factors affecting microplastic ingestion by
D. polymorpha
Bivalves selectively filter‐feed on particles from the water
column. The mussels' ciliated gills as well as ciliated labial palps
separate digestible food particles from the remaining particle
load by transporting the former to the digestive system and
rejecting the latter as pseudofeces (Ward and Shumway 2004;
Vaughn et al. 2008; Tuttle‐Raycraft and Ackerman 2019). This
selection mechanism is, however, inefficient for microplastics
because we detected plastic particles in almost every individual
after a respective exposure. Accordingly, more knowledge on
the factors affecting the ingestion is required to better under-
stand the mussels' exposure to microplastics.
Exposure and depuration time. Dreissena polymorpha
rapidly ingested microplastic, with a maximum body burden
reached after 1 and 12 h (experiment 1). The excretion of mi-
croplastic was similarly fast in D. polymorpha, with a significant
FIGURE 4: Chronic toxicity of polystyrene microplastic fragments (≤63 µm, 6.4–100 000 p mL
–1
) and diatomite (≤63 µm, 100 000 p mL
–1
)in
Dreissena polymorpha exposed for 1, 3, 7, and 42 d. Endpoints were (A) protein, (B) glycogen, (C) total lipids, (D) malondialdehyde content, and (E)
the remaining antioxidant capacity (Trolox equivalents) in the midgut gland, as well as (F) clearance rate. Dots indicate mean of each treatment; lines
indicate linear regressions (separate regression for each exposure duration). (A–E)n=7–10, (F)n=10. MGG =midgut gland. MDA =
malondialdehyde; EQ =equivalent.
Polystyrene microplastics in freshwater bivalves—Environmental Toxicology and Chemistry, 2021;40:2247–2260 2255
wileyonlinelibrary.com/ETC © 2021 The Authors

reduction of microplastic body burden after 1 h of depuration
and a reduction to fewer than 10 particles from 3 h onward.
These results agree well with data on marine bivalves; for ex-
ample, Mytilus edulis ingests polyethylene terephthalate fibers
with a maximum body burden after 6 h (Woods et al. 2018), and
Geukensia demissa excretes ingested polyethylene spheres
within 12 h (Khan and Prezant 2018). This confirms that maximal
microplastic burden in mussels is reached already after a few
hours of exposure. Furthermore, decreasing microplastic bur-
dens following peak levels highlight the fast egestion of mi-
croplastic particles by mussels but possibly also indicate
decreasing exposure concentrations throughout prolonged
exposure periods.
Despite fast excretion, few microplastics remained in
D. polymorpha after 7 d of depuration. Therefore, mussel
clearance mechanisms do not eliminate all internalized mi-
croplastics within 1 wk after exposure. The residual particles
mayhavebeeneitherretainedinthedigestivetractor
translocated into tissues or the circulatory system. In previous
research, microplastics (≤25 µm) were detected not only in the
stomach and intestine but also in the associated midgut gland
ducts and diverticula (blind‐ending tubules [Owen 1974]) and
the circulatory system of bivalves (Von Moos et al. 2012;
Guilhermino et al. 2018; Magni et al. 2018; Pittura et al. 2018;
Gonçalves et al. 2019). These translocation processes may
prevent the removal of the microplastic particles from the
body and, thus, cause prolonged retention in mussels such as
D. polymorpha.
When considering microplastic ingestion after extended
time periods (>12 h), we observed a constant decrease of the
microplastic body burden without reaching a steady state. This
indicates that the microplastic exposure or bioavailability de-
creased during the experiment, possibly because the micro-
plastic sedimented or the mussels had taken up the available
microplastic and excreted it as (pseudo)feces. We did not ex-
amine feces production and deposition in our study and,
therefore, cannot quantify its contribution to bioavailability
reduction. Nonetheless, we investigated the settlement of
particles in an additional experiment (Supplemental Data, S2.2)
and observed that from 6 h onward, only 5‐and 10‐µm spheres
remained in the water phase and that after 12 h only approx-
imately half the 5‐µm spheres and one‐third of the 10‐µm
spheres remained in the water phase. This indicates that from
12 h onward, microplastic uptake by D. polymorpha was pos-
sible lower over time, whereas a constant depuration resulted
in an overall decreasing microplastic body burden. From this,
we conclude that stable microplastic burden in mussels from
wild populations are only possible in case of continuous mi-
croplastic bioavailability and ingestion by the mussels.
Body size. The absolute microplastic ingestion (per
individual) varied distinctively within each size class of
D. polymorpha. Accordingly, no significant differences be-
tween the 3 size classes were observed. This indicates that the
interindividual differences in absolute microplastic burden in
mussels may be unrelated to body size. Considering the rela-
tive body burden (microplastic per dry wt), instead, microplastic
ingestion was significantly higher in smaller individuals. Hence,
smaller mussels (e.g., juveniles) in the environment may be
exposed to relatively higher microplastic levels. This is possibly
related to higher relative feeding activity. In Mytilus spp., the
weight‐specific pumping rate increases exponentially with de-
creasing body size. The same is true for the weight‐specific gill
area (Jones et al. 1992; Duinker et al. 2007). Higher relative
pumping rates and larger relative gill areas allow smaller
mussels to take up more microplastics per body mass.
Food abundance. Providing algae as food caused a sig-
nificant, dose‐dependent decrease in microplastic ingestion in
D. polymorpha (experiment 3). Rist et al. (2019a) observed the
same trend for M. edulis larvae, which were exposed to 2‐µm
PS spheres in the absence and presence of algae. In adult
M. edulis instead, higher algae concentrations increased the
ingestion of 30‐nm PS spheres (Wegner et al. 2012). Wegner
et al. (2012) discuss that nanospheres may adsorb on the
algae as a potential reason for their findings. Because we used
microplastics with a similar or larger size than D. subspicatus
cells (8 ×5µm; Hessen and Van Donk 1993), indirect micro-
plastic ingestion through adsorption on algae was probably
not a relevant pathway. Considering environmental con-
ditions, these results indicate that high food abundance may
reduce overall microplastic ingestion by bivalves. Especially in
freshwater systems with high primary production or in seasons
with elevated food availability (e.g., algae bloom in spring),
TABLE 1: Two‐way analysis of variance results for the effects of the variables “microplastic concentration”and “exposure time”and their inter-
action on the clearance rate, energy reserves (proteins, glycogen, total lipids), and oxidative stress (TBARS, ORAC) of Dreissena polymorpha
Variable Clearance rate Proteins Glycogen Total lipids TBARS ORAC
Microplastic concentration df 444444
F3.205 2.099 0.483 0.918 2.132 1.545
p0.014 0.083 0.748 0.455 0.079 0.192
Exposure time df 333333
F0.861 7.671 5.320 6.348 4.670 11.254
p0.463 <0.001 0.002 <0.001 0.004 <0.001
Microplastic concentration ×exposure time df 12 12 12 12 12 12
F0.990 1.117 1.680 1.356 1.133 0.910
p0.460 0.350 0.075 0.192 0.337 0.538
Bold indicates significance.
TBARS =thiobarbituric acid reactive substances; ORAC =oxygen radical absorbance capacity; MP =microplastic; DI =diatomite.
2256 Environmental Toxicology and Chemistry, 2021;40:2247–2260—A. Weber et al.
© 2021 The Authors wileyonlinelibrary.com/ETC

mussels may take up fewer microplastics than in oligotrophic
systems.
Microplastic concentration. Dreissena polymorpha in-
gested microplastics in a concentration‐dependent manner
(experiment 4). The increase in microplastic uptake was, how-
ever, not linear such that animals exposed to a 10‐fold higher
concentration had 10‐fold higher body burdens. Therefore,
microplastic body burdens in mussels may behave rather log-
arithmically, suggesting the existence of a maximal level. A
logarithmic relation between microplastic water concentrations
and body burdens has previously been reported for M. gallo-
provincialis larvae (Capolupo et al. 2018) and M. edulis
(Woods et al. 2018). Limited microplastic uptake at very high
concentrations in the water phase is mostly related to
pseudofeces production. When the gill cilia which sort particles
prior to ingestion (see section Ingestion of microplastics in
3 mussel species) are overloaded, excess particles are gath-
ered, embedded into pseudofeces, and rejected (Ward et al.
1993). However, overloading of the bivalves' sorting mecha-
nism with microplastics is possibly not very relevant in nature
because the highest reported environmental concentrations
of 0.2 to 0.5 p mL
–1
(Leslie et al. 2017; Lahens et al. 2018)
are much lower than the concentrations used in the
laboratory studies mentioned above (Capolupo et al. 2018
[50–10 000 p mL
–1
]; Woods et al. 2018 [3–30 p mL
–1
]).
Ingestion of microplastics in 3 mussel species
The ingestion experiments with D. polymorpha pointed out
that numerous factors affect microplastic ingestion in this
species, including exposure and depuration time, body size,
food abundance, and microplastic concentration. Species‐
specific differences (e.g., size, morphology) may, however, be
another relevant factor influencing microplastic ingestion by
freshwater mussels. In light of a broader assessment of micro-
plastic ingestion in freshwater species, we repeated the ex-
periments for the factors exposure and depuration time
(experiment 1) as well as body size (experiment 2) with the
2 larger freshwater mussel species, A. anatina and S. woodiana.
Distinct differences between the 3 freshwater species were
seen with regard to the absolute microplastic ingestion. The
smaller D. polymorpha ingested fewer microplastics than the
2 larger species (experiment 1) when considering the absolute
body burden. Between the 2 larger species instead, absolute
microplastic ingestion did not differ distinctively. Higher mi-
croplastic ingestion in larger species may be caused by either
higher absolute filtration rates of larger mussels as reported by
Kryger and Riisgård (1988) or a higher bioavailability of mi-
croplastics to the larger, sand‐dwelling unionids. The latter
would be true if the sand‐dwelling species are able to take up
additional microplastics from the sediment. Based on an ad-
ditional experiment (see Supplemental Data, S6), we did,
however, show that microplastic uptake by unionids from the
sediment phase is rather limited and that the filtration rate is,
thus, the determining factor for absolute microplastic in-
gestion. Accordingly, body size seems to be a relevant
indicator for microplastic body burden, with larger species in-
gesting higher microplastic quantities. Differences between
species of similar size, instead, seem to be of limited relevance
for absolute microplastic ingestion.
In contrast, the relative microplastic body burden (per dry wt)
was higher in smaller specimens, with both species and individual
size having a significant effect. Variations in relative microplastic
ingestion are possibly caused by differences in relative filtration
rate between mussels of different species and size. Kryger and
Riisgård (1988) observed that smaller bivalves (e.g., D. poly-
morpha) often have a higher relative filtration rate (per mussel dry
wt) compared to larger species (e.g., Unionidae) but that filtration
rates may also vary between species of similar size (e.g.,
Sphaerum corneum vs D. polymorpha). This suggests that smaller
species may be more susceptible to microplastic ingestion than
larger ones but that species‐specific variations may also affect the
overall microplastic exposure in mussel individuals.
Further, we compared the ingestion and depuration be-
havior of D. polymorpha and S. woodiana over various ex-
posure and depuration time periods. Ingestion behavior over a
period of up to 12 h was rather similar because both species
reached peak levels within 12 h. Depuration, instead, differed
in both species, with D. polymorpha excreting microplastics
faster than S. woodiana.Dreissena polymorpha removed in-
gested PS spheres almost completely within 12 h, whereas
microplastic levels in S. woodiana decreased further until 72 h
of depuration. Because we compared 2 species with different
size, it remains unclear whether the difference is species‐or
rather size‐specific. The latter has already been shown for
Mytilus chilensis, in which smaller individuals had a higher
weight‐specific excretion rate compared to larger ones (Nav-
arro and Winter 1982). Hence, smaller mussels have a higher
relative microplastic body burden but also excrete micro-
plastics faster than larger mussels. In smaller mussels, higher
relative uptake may, therefore, be compensated for by a higher
microplastic depuration.
Interestingly, the 3 freshwater species ingested microplastic
size‐dependently. Although 10‐µm PS spheres were most abun-
dant in all species, D. polymorpha also ingested large quantities
of 5‐µm spheres, whereas the larger species tended to ingest
larger particles (45‐and 90‐µm spheres). Variations in feeding size
selectivity are possibly due to differences in the particle selection
mechanism and, more specifically, to morphological variations of
the lamellated gills (Ward and Shumway 2004; Rosa et al. 2018),
as well as the selection mechanism in the stomach. Lamellated
gills are composed of numerous filaments with associated later-
ofrontal cirri (bundle of cilia) with which inhaled particles are se-
lected from the inhaled water current (Ward and Shumway 2004;
Silverman et al. 1999). Jørgensen et al. (1984) have shown that
the retention potential of the laterofrontal cirri is species‐
dependent. Although D. polymorpha efficiently retains particles
as small as 1.5 µm with the laterofrontal cirri, in Anodonta cygnea
4‐µm particles pass through the cirri. The difference in latero-
frontal cirri morphology may, thus, be a reason for higher in-
gestion of 5‐µmspheresinD. polymorpha.
Following the lamellated gills, particle selection involves
further sorting steps on the labial palps as well as within the
Polystyrene microplastics in freshwater bivalves—Environmental Toxicology and Chemistry, 2021;40:2247–2260 2257
wileyonlinelibrary.com/ETC © 2021 The Authors

digestive system. As a final step, particles are separated in
specialized grooves in the stomach and either redirected into
the digestive tubules of the midgut gland or into the intestine
for direct excretion (Ward and Shumway 2004). Ten Winkel and
Davids (1982) reported that the stomach of D. polymorpha
preferentially selects for algae with a size between 10 and
50 µm, with a maximal preference for 20‐µm algae cells. This
size preference was also seen in our ingestion experiments
because all tested freshwater species mostly ingested 10‐µm
spheres. Still, a high burden of 5‐µm PS spheres, especially in
D. polymorpha, indicates that stomach sorting is not exclusively
limited to 10‐to 50‐µm particles.
Besides interspecies differences, we also observed that the
size selectivity varied in each species. For all 3 tested species,
smaller individuals ingested larger quantities of 5‐and 10‐µm
spheres compared to larger individuals (experiment 2). Again, this
is probably related to morphological changes once a bivalve
grows: for M. edulis, the gill and labial palp area as well as the
distance between gill filaments enlarge with increasing body
length (Kiørboe and Møhlenberg 1981; Jones et al. 1992), and
larger interfilamentary spaces will reduce the retention of small
particles. This points toward a higher relevance of an ingestion of
small microplastics for smaller individuals (e.g., juveniles).
Microplastic toxicity in D. polymorpha
The exposure of D. polymorpha to PS fragments (≤63 µm)
for 1 to 42 d at concentrations up to 100 000 p mL
–1
caused a
significant increase in clearance rate, whereas mortality, met-
abolic endpoints (energy reserves), and oxidative stress were
not altered. Hence, the increase in clearance rate, especially in
the 100 000 p mL
–1
microplastic treatment, may be an adapta-
tion mechanism to compensate for potential microplastic‐
related effects (e.g., reduced food uptake). Considering the
exposure time, we observed a significant general change in
energy allocation as well as basic stress levels in D. polymorpha
throughout the 42‐d exposure. Because D. polymorpha in-
dividuals were accustomed to the medium as well as the food
source as early as 4 wk prior to the toxicity study, we assume
that changes in endpoints over the exposure time were not
caused by the exposure design itself but may rather be related
to stress reactions caused by the water changes and the
transfer of the mussels into new exposure tanks. Such a stress
reaction could have, thus, also caused the observed decrease
of the antioxidative capacity with increasing exposure time. The
definite reason, however, remains unclear.
Interestingly, we also did not observe significant differences
between the effects of the microplastic and diatomite treat-
ments. This suggests that D. polymorpha is rather insensitive to
an exposure to high concentrations of microplastic and natural
particles alike. These results contrast with those of earlier
studies in which mussels reduced their clearance rate with in-
creasing turbidity, resulting in a lower food uptake (Aldridge
et al. 1987; Tuttle‐Raycraft and Ackerman 2019). Mussels are
able to compensate for such reduced feeding by increasing the
particle selection efficiency via morphological adaptations
(Payne et al. 1995; Tuttle‐Raycraft and Ackerman 2019). Hence,
compensation mechanisms to microplastic exposure (increase
in clearance rate) in our study seem to differ from those for high
turbidity. It will, however, require additional studies to eluci-
date whether these differences are stressor‐dependent (mi-
croplastic vs natural particle turbidity) or a result of different
adaptation periods (42‐d microplastic exposure vs lifelong ex-
posure in turbid environment). Notwithstanding, compensation
mechanisms seemed to be efficient enough to protect D.
polymorpha from microplastic effects even at concentrations
far higher than currently reported for freshwater environments.
The absence of negative effects in our study only partially
agrees with previous research. In a systematic literature search
in May 2020 (PubMed, search “microplastics”OR “micro-
plastic”AND “mussel”OR “bivalve,”studies evaluating tox-
icity after depuration excluded), we identified 8 toxicity studies
with freshwater mussels (D. polymorpha and C. fluminea; see
Supplemental Data, Table S5). Two studies did not report any
microplastic‐induced effects, whereas one study observed ef-
fects on all and 5 studies on some of the analyzed endpoints.
Although the majority of the endpoints remained unaffected
(Supplemental Data, Table S5), these reports show that fresh-
water bivalves may be, at least to some extent, susceptible to
microplastic exposure. However, applied microplastic concen-
trations in the toxicity studies with D. polymorpha and C.
fluminea exceeded those currently reported for freshwater
ecosystems (0.52 p mL
–1
[Lahens et al. 2018]). It, thus, remains
unknown whether the observed effects are relevant in the en-
vironmental context.
Interestingly, effects on feeding, histology, neurology, and
oxidative stress metabolites varied intensively, with similar end-
points being affected in some but not in other studies. Effects
may be species‐specific; however, even when comparing the
same species (D. polymorpha), marked differences were ob-
served (e.g., effects on catalase and glutathione peroxidase in
Magni et al. [2018, 2020]). We, therefore, believe that complex
effect patterns are also a consequence of differences in exposure
scenarios (especially exposure time and concentration), micro-
plastic properties (e.g., polymer type, size, shape), and the
sensitivity of endpoints. Accordingly, drawing general con-
clusions on microplastic risks for bivalves may be overly sim-
plistic. An identification of the most sensitive species and their
respective traits provides a way forward. Beyond that, a better
understanding of the biological mechanisms protecting mussels
from high loads of suspended particles as well as of the mech-
anism of microplastic toxicity in mussels is required. This
knowledge is fundamental to determining whether microplastics
can indeed bypass defense mechanisms in bivalves and, thus,
represents a risk to wild mussel populations.
CONCLUSION
We investigated the kinetics of microplastic ingestion and
egestion in D. polymorpha. Comparing multiple relevant factors,
we demonstrate that exposure and depuration time, body size,
food abundance, and microplastic concentration affect overall
microplastic ingestion. Dreissena polymorpha rapidly ingested
microplastics and excreted the majority of the particles within
2258 Environmental Toxicology and Chemistry, 2021;40:2247–2260—A. Weber et al.
© 2021 The Authors wileyonlinelibrary.com/ETC

12 h, with only a few particles being retained for more than 1 wk.
Smaller individuals ingested more microplastics compared to
larger individuals relative to their body size. An additional supply
of food reduced the uptake of microplastics.
Further, we compared microplastic ingestion in D. poly-
morpha with 2 larger unionid species (A. anatina,S. woodiana).
Absolute microplastic ingestion was higher in the 2 unionid
species but did not differ between the unionids. Relative mi-
croplastic ingestion, instead, differed significantly with regard
to both the species and the body size of the individuals, with
smaller mussels ingesting more microplastics. With regard to
microplastic size, smaller species as well as smaller individuals
within each species also ingested smaller microplastics.
We also analyzed the toxicity of PS fragments (≤63 µm,
6.4–100 000 p mL
–1
)inD. polymorpha. Exposure for up to 42 d
caused a significant increase in the clearance rate but did not
affect energy reserves or oxidative stress. Further, no sig-
nificant difference between the effects of microplastics and
natural particles (diatomite) was observed. Enhanced filtration
rate may be a compensatory mechanism rendering D. poly-
morpha rather insensitive to microplastic exposure even at very
high concentrations. Taking into account previous research on
microplastic toxicity in bivalves, this does not imply that mus-
sels in general are not susceptible to microplastic exposures.
Divergent toxicity data probably originate from species‐specific
differences or variations in experimental design. A better un-
derstanding of the traits and mechanisms rendering some
species and endpoints more sensitive than others is needed to
prioritize species potentially at risk.
Supplemental Data—The Supplemental Data are available on
the Wiley Online Library at https://doi.org/10.1002/etc.5076.
Acknowledgment—We thank J. Oehlmann for his fruitful
feedback on the experiments and for providing the laboratory
infrastructure. Further, we thank the ASV Walldorf 1962 e.V. for
providing access to the D. polymorpha sampling site. The
present study was funded by the German Federal Ministry of
Transport and Digital Infrastructure.
Author Contributions Statement—A. Weber conceived the
study; A. Weber, N. Jeckel, C. Weil, and S. Umbach performed
the study; A. Weber, N. Jeckel, and C. Weil analyzed the data;
M. Wagner provided feedback on the study results; A. Weber
and M. Wagner wrote the manuscript; all authors commented
on the manuscript.
Data Availability Statement—Data, associated metadata, and
calculation tools are available from the corresponding author
(martin.wagner@ntnu.no).
REFERENCES
Aldridge D, Payne B, Miller A. 1987. The effects of intermittent exposure to
suspended solids and turbulence on three species of freshwater
mussels. Environ Pollut 45:17–28.
Baudrimont M, Arini A, Guégan C, Venel Z, Gigault J, Pedrono B, Prunier J,
Maurice L, Ter Halle A, Feurtet‐Mazel A. 2020. Ecotoxicity of poly-
ethylene nanoplastics from the North Atlantic oceanic gyre on fresh-
water and marine organisms (microalgae and filter‐feeding bivalves).
Environ Sci Pollut Res Int 27:3746–3755.
Benedict M. 2014. Methods in Anopheles Research. BEI Resources, Mana-
ssas, VA, USA. [cited 2021 April 11]. Available from: https://www.
beiresources.org/Publications/MethodsinAnophelesResearch.aspx
Binelli A, Pietrelli L, Di Vito S, Coscia L, Sighicelli M, Della Torre C, Parenti C,
Magni S. 2020. Hazard evaluation of plastic mixtures from four Italian
subalpine great lakes on the basis of laboratory exposures of zebra
mussels. Sci Total Environ 699:134366.
Bradford M. 1976. A rapid and sensitive method for the quantitation of
microgram quantities of protein utilizing the principle of protein‐dye
binding. Anal Biochem 72:248–254.
Bråte I, Blázquez M, Brooks S, Thomas K. 2018. Weathering impacts the
uptake of polyethylene microparticles from toothpaste in Mediterranean
mussels (M. galloprovincialis). Sci Total Environ 626:1310–1318.
Brillant M, MacDonald B. 2000. Postingestive selection in the sea scallop,
Placopecten magellanicus (Gmelin): The role of particle size and density.
J Exp Mar Biol Ecol 253:211–227.
Capolupo M, Franzellitti S, Valbonesi P, Lanzas CS, Fabbri E. 2018. Uptake
and transcriptional effects of polystyrene microplastics in larval stages of
the Mediterranean mussel Mytilus galloprovincialis.Environ Pollut
241:1038–1047.
De Sá L, Oliveira M, Ribeiro F, Rocha T, Futter M. 2018. Studies of the
effects of microplastics on aquatic organisms: What do we know and
where should we focus our efforts in the future? Sci Total Environ
645:1029–1039.
Duinker A, Bergslien M, Strand Ø, Olseng C, Svardal A. 2007. The effect of
size and age on depuration rates of diarrhetic shellfish toxins (DST) in
mussels (Mytilus edulis L.). Harmful Algae 6:288–300.
Fernández B, Albentosa M. 2019. Insights into the uptake, elimination
and accumulation of microplastics in mussel. Environ Pollut
249:321–329.
Furuhagen S, Liewenborg B, Breitholz M, Gorokhova E. 2014. Feeding ac-
tivity and xenobiotics modulate oxidative status in Daphnia magna:
Implications for ecotoxicological testing. Environ Sci Technol
48:12886–12892.
Gonçalves C, Martins M, Sobral P, Costa P, Costa M. 2019. An assessment
of the ability to ingest and excrete microplastics by filter‐feeders:
A case study with the Mediterranean mussel. Environ Pollut
245:600–606.
Guilhermino L, Vieira L, Ribeiro D, Tavares A, Cardoso V, Alves A, Almeida
J. 2018. Uptake and effects of the antimicrobial florfenicol, microplastics
and their mixtures on freshwater exotic invasive bivalve Corbicula
fluminea.Sci Total Environ 622–623:1131–1142.
Hartmann N, Hüffer T, Thompson R, Hassellöv M, Verschoor A, Daugaard A,
Rist S, Karlsson T, Brennholt N, Cole M, Herrling M, Hess C, Ivleva N,
Lusher A, Wagner M. 2019. Are we speaking the same language?
Recommendations for a definition and categorization framework for
plastic debris. Environ Sci Technol 53:1039–1047.
Hessen D, Van, Donk E. 1993. Morphological changes in Scenedesmus
induced by substances released from Daphnia.Arch Hydrobiol
127:129–140.
Hodges D, DeLong J, Forney C, Prange R. 1999. Improving the thio-
barbituric acid‐reactive‐substances assay for estimating lipid perox-
idation in plant tissues containing anthocyanin and other interfering
compounds. Planta 207:604–611.
Jones H, Richards O, Southern T. 1992. Gill dimensions, water pumping rate
and body size in the mussel Mytilus edulis L. J Exp Mar Biol Ecol
155:213–237.
Jørgensen C, Kiørboe T, Møhlenberg F, Riisgård H. 1984. Ciliary and
mucus‐net filter feeding, with special reference to fluid mechanical
characteristics. Mar Ecol Prog Ser 15:283–292.
Khan M, Prezant R. 2018. Microplastic abundances in a mussel bed and
ingestion by the ribbed marsh mussel Geukensia demissa.Mar Pollut
Bull 130:67–75.
Kiørboe T, Møhlenberg F. 1981. Particle selection in suspension‐feeding
bivalves. Mar Ecol Prog Ser 5:291–296.
Kryger J, Riisgård H. 1988. Filtration rate capacities in 6 species of European
freshwater bivalves. Oecologia 77:34–38.
Polystyrene microplastics in freshwater bivalves—Environmental Toxicology and Chemistry, 2021;40:2247–2260 2259
wileyonlinelibrary.com/ETC © 2021 The Authors

Lahens L, Strady E, Kieu‐Le TC, Dris R, Boukerma K, Rinnert E, Gasperi J,
Tassin B. 2018. Macroplastic and microplastic contamination assessment
of a tropical river (Saigon River, Vietnam) transversed by a developing
megacity. Environ Pollut 236:661–671.
Leslie H, Brandsma S, van Velzen M, Vethaak A. 2017. Microplastic en route:
Field measurements in the Dutch river delta and Amsterdam canals,
wastewater treatment plants, North Sea sediments and biota. Environ Int
101:133–142.
Li J, Lusher A, Rotchell J, Deudero S, Turra A, Bråte I, Sun C, Hossain M, Li
Q, Kolandhasamy P, Shi H. 2019. Using mussel as a global bioindicator
of coastal microplastic pollution. Environ Pollut 244:522–533.
Li L, Su L, Cai H, Rochman C, Li Q, Kolandhasamy P, Peng J, Shi H. 2019.
The uptake of microfibers by freshwater Asian clams (Corbicula fluminea)
varies based upon physiochemical properties. Chemosphere
221:107–114.
Magni S, Bonasoro F, Della Torre C, Parenti C, Maggioni D, Binelli A. 2020.
Plastics and biodegradable plastics. Ecotoxicity comparison between
polyvinylchloride and Mater‐Bi
®
micro‐debris in a freshwater biological
model. Sci Total Environ 720:137602.
Magni S, Della Torre C, Garrone G, D'Amato A, Parenti C, Binelli A. 2019.
First evidence of protein modulation by polystyrene microplastics in a
freshwater biological model. Environ Pollut 250:407–415.
Magni S, Gagné F, André C, Della Torre C, Auclair J, Hanana H, Parenti C,
Bonasoro F, Binelli A. 2018. Evaluation of uptake and chronic toxicity of
virgin polystyrene microbeads in freshwater zebra mussel Dreissena
polymorpha (Mollusca: Bivalvia). Sci Total Environ 631–632:778–788.
Mathalon A, Hill P. 2014. Microplastic fibers in the intertidal ecosystem
surrounding Halifax Harbor, Nova Scotia. Mar Pollut Bull 81:69–79.
Mercogliano R, Avio C, Regoli F, Anastasio A, Colavita G, Santonicola S.
2020. Occurrence of microplastics in commercial seafood under the
perspective of the human food chain. A review. J Agric Food Chem
68:5296–5301.
Navarro J, Winter J. 1982. Ingestion rate, assimilation efficiency and energy
balance in Mytilus chilensis in relation to body size and different algal
concentrations. Mar Biol 67:255–266.
Oliveira P, Barboza L, Branco V, Figueiredo N, Carvalho C, Guilhermino L.
2018. Effects of microplastics and mercury in the freshwater bivalve
Corbicula fluminea (Müller, 1774): Filtration rate, biochemical bio-
markers and mercury bioconcentration. Ecotoxicol Environ Saf
164:155–163.
Organisation for Economic Co‐operation and Development. 2016. Test No.
242: Potamopyrgus antipodarum reproduction test. OECD Guidelines
for the Testing of Chemicals. Paris, France.
Ou B, Hampsch‐Woodill M, Prior R. 2001. Development and validation of an
improved oxygen radical absorbance capacity assay using fluorescein as
the references fluorescent probe. J Agric Food Chem 49:4619–4626.
Owen G. 1974. Feeding and digestion in the Bivalvia. Adv Comp Physiol
Biochem 5:1–35.
Payne B, Lei J, Miller A, Hubertz E. 1995. Adaptive variation in palp and gill
size of the zebra mussel (Dreissena polymorpha) and Asian clam
(Corbicula fluminea). Can J Fish Aquat Sci 52:1130–1134.
Péden R, Poupin P, Sohm B, Flayac J, Giambérini L, Klopp C, Louis F,
Pain‐Devin S, Potet M, Serre RF, Devin S. 2019. Environmental tran-
scriptomes of invasive dreissena, a model species in ecotoxicology and
invasion biology. Sci Data 6:234.
Pittura L, Avio CG, Giuliani M, d'Errico G, Keiter S, Cormier B, Gorbi S,
Regoli F. 2018. Microplastics as vehicles of environmental PAHs to
marine organisms: Combined chemical and physical hazards to the
Mediterranean mussels, Mytilus galloprovincialis. Front Mar Sci 5:103.
PlasticsEurope. 2018. Plastics, the facts 2018. An analysis of European
plastics production, demand and waste data. Brussels, Belgium. [cited
2021 April 11]. Available from: https://www.plasticseurope.org/
application/files/6315/4510/9658/Plastics_the_facts_2018_AF_web.pdf
Railo S, Talvitie J, Setälä O, Koistinen A, Lehtiniemi M. 2018. Application of
an enzyme digestion method reveals microliter in Mytilus trossulus at a
wastewater discharge area. Mar Pollut Bull 130:206–214.
Rist S, Baun A, Almeda R, Hartmann N. 2019a. Ingestion and effects of
micro‐and nanoplastics in blue mussel (Mytilus edulis) larvae. Mar Pollut
Bull 140:423–430.
Rist S, Steensgaards I, Guven O, Nielsen T, Jensen L, Møller L, Hartmann N.
2019b. The fate of microplastics during uptake and depuration phases in
a blue mussel exposure system. Environ Toxicol Chem 38:99–105.
Rochman C, Parnis J, Browne M, Serrato S, Reiner E, Robson M, Young T,
Diamond M, Teh S. 2017. Direct and indirect effects of different types of
microplastics on freshwater prey (Corbicula fluminea) and their predator
(Acipenser transmontanus). PLoS One 12:e0187664.
Rosa M, Ward J, Shumway S. 2018. Selective capture and ingestion of
particles by suspension‐feeding bivalve molluscs: A review. J Shellfish
Res 37:727–746.
Setälä O, Norkko J, Lehtiniemi M. 2016. Feeding type affects microplastic
ingestion in a coastal invertebrate community. Mar Pollut Bull
102:95–101.
Silverman H, Lynn J, Beninger P, Dietz T. 1999. The role of latero‐frontal cirri
in particle capture by the gills of Mytilus edulis.Biol Bull 197:368–376.
Ten Winkel E, Davids C. 1982. Food selection by Dreissena polymorpha
Pallas (Mollusca: Bivalvia). Freshw Biol 12:553–558.
Tuttle‐Raycraft S, Ackerman J. 2019. Living the high turbidity life: The ef-
fects of total suspended solids, flow, and gill morphology on mussel
feeding. Limnol Oceanogr 64:2526–2537.
Vaughn C, Nichols S, Spooner D. 2008. Community and foodweb ecology
of freshwater mussels. J North Am Benthol Soc 27:409–423.
Von Moos N, Burkhardt‐Holm P, Köhler A. 2012. Uptake and effects of
microplastics on cells and tissue of the blue mussel Mytilus edulis L. after
an experimental exposure. Environ Sci Technol 46:11327–11335.
Ward J, MacDonald B, Thompson R. 1993. Mechanisms of suspension
feeding in bivalves: Resolution of current controversies by means of
endoscopy. Limnol Oceanogr 38:265–272.
Ward J, Shumway S. 2004. Separating the grain from the chaff: Particle
selection in suspension‐and deposit‐feeding bivalves. J Exp Mar Biol
Ecol 300:83–130.
Weber A, Jeckel N, Wagner M. 2020. Combined effects of polystyrene
microplastics and thermal stress on the freshwater mussel Dreissena
polymorpha.Sci Total Environ 718:137253.
Weber A, von Randow M, Voigt AL, von der Au M, Fischer E, Meermann B,
Wagner M. 2021. Ingestion and toxicity of microplastics in the fresh-
water gastropod Lymnaea stagnalis: No microplastic‐induced effects
alone or in combination with copper. Chemosphere 263:128040.
Wegner A, Besseling E, Foekema E, Kamermans P, Koelmans A. 2012. Ef-
fects of nanopolystyrene on the feeding behavior of the blue mussel
(Mytilus edulis L.). Environ Toxicol Chem 31:2490–2497.
Woods M, Stack M, Fields D, Shaw S, Matrai P. 2018. Microplastic fiber
uptake, ingestion, and egestion rates in the blue mussel (Mytilus edulis).
Mar Pollut Bull 137:638–645.
2260 Environmental Toxicology and Chemistry, 2021;40:2247–2260—A. Weber et al.
© 2021 The Authors wileyonlinelibrary.com/ETC