Content uploaded by Thomas Kirk Sørensen
Author content
All content in this area was uploaded by Thomas Kirk Sørensen on May 09, 2016
Content may be subject to copyright.
Analysis of microplastic in the
stomachs of herring and cod from
the North Sea and Baltic Sea
Robin Lenz, Kristina Enders, Sabrina Beer, Thomas Kirk
Sørensen, Colin A. Stedmon
Photos: Line Reeh, Robin Lenz, Kristina Enders
Summary
There is a pressing need to assess the distribution of microplastic in aquatic
environments and the extent to which they are ingested by fish. This report
presents the results of a study contracted by the Danish Ministry of Environment
and Food’s Nature Agency in 2015 to analyse the microplastic stomach content
of demersal and pelagic fish from the North Sea and Baltic Sea. The focus is on
particles > 100 µm in size and on comparing distributions in coastal and offshore
waters using the sampling already planned as part of DTU Aqua fish monitoring
activities. Overall 23% of the analysed fish contained one or more particles or
fibres of synthetic polymeric material. Although cod contained plastic more often
and also in higher numbers than herring, the latter species showed higher
numbers of microplastic items per gram of stomach analysed. A small subsample
of the retrieved microplastics was analysed under a Raman microspectrometer
revealing common commodity plastic polymer types such as PE, PP and PS, but
no quantitative analysis of these was possible within the project’s scope. The
comparison between offshore and coastal regions is hindered by the fact that
many of the coastal stations were close to the outer boundary of the defined
coastal zone (12 nautical miles from shore) due to sampling from commercial
fisheries and large research vessels.
We recommend the development of an indicator for microplastic in fish based on
the present study, where stable coastal populations in close proximity to urban
areas could be compared against a reference group unaffected by direct land-
based microplastic input.
Introduction
The European Union Marine Strategy Framework Directive lists eleven
qualitative descriptors for determining good environmental status (Anon 2008).
One of these (Descriptor 10) is focused on marine litter and states that, in order
to achieve Good Environmental Status regarding marine litter, member states
must ensure that: “Properties and quantities of marine litter do not cause harm
to the coastal and marine environment”. In this light member states now have to
begin to assess the distribution and impact of solid debris such as plastic. In 2010
the European Commission further specified four indicators for descriptor 10,
where identifying “trends in the amount and composition of litter ingested by
marine animals (e.g. stomach analysis)” is one (Galgani and Hanke 2013).
Good environmental status for marine litter is in the Danish marine strategy
(NST 2012) described as: 1) litter and its degradation products do not cause
harm to marine ecosystems and species and do not support spreading of non-
indigenous and invasive species; and 2) litter and its degradation products do
not have a significant negative socio-economic impact on marine professions and
professions associated with marine areas including tourism. Furthermore three
environmental targets have been set, which describe intermediate goals towards
reaching good environmental status. Due to lack of knowledge, an operational
target has not yet been formulated for microparticles. The Danish marine
strategy states that, in order to develop quantitative targets, scientific data is
needed to establish reference levels and to specify actions to achieve significant
1
reductions in litter, including microplastic particles. To inform the development
of indicators to support target-setting, Denmark’s MSFD monitoring strategy
(NST 2014) includes monitoring of microplastic in fish. The current study is
part of this monitoring programme and is prescribed in (NST 2014) as follows:
Macro- and microliter in fish stomachs as prescribed in EU Guideline ”Guidance
on Monitoring of Marine Litter in European Seas” (TSG-ML 2014). It is
suggested that 2 fish species are included with different feeding strategies, so
litter in both water column and seafloor is represented. The chemical
composition of microlitter (< 5mm) is determined spectroscopically to gather
information on possible sources of microlitter. It is suggested to monitor litter in
fish stomachs at least once in the programme period.
Awareness of plastic waste in the sea has been rapidly increasing in recent years,
in particular the widespread occurrence of microplastic (Ivar do Sul and Costa
2014). Although macroplastic litter from human activity in the form of bags,
containers, ropes etc. were known to persist and be transported over large
distances, the realisation of a globally widespread distribution of microplastic
particles has obtained much media attention. Microplastic consists of particles
less than 5 mm in size, that in part originate from the disintegration of
macroplastic in the environment (Barnes et al. 2009) but are also specifically
manufactured and eventually released via, for example, waste water effluent
(Gregory 1996; Fendall and Sewell 2009).
Concerns on the impact of plastic litter ingestion on marine organisms can be
grouped into two categories: physical congestion and damage of digestive tracts
and the role of microplastic as adsorbent surfaces for pollutants. The size range
of microplastic particles overlaps with that of many planktonic organisms and as
a result they are commonly ingested by detritivores and planktivores (Wright,
Thompson, and Galloway 2013). Microplastic can be either directly ingested by
fish or indirectly through feeding on zooplankton which have ingested
microplastic (Cole et al. 2013; Setälä, Fleming-Lehtinen, and Lehtiniemi 2014).
Additionally the surface of plastic particles efficiently scavenges hydrophobic
persistent pollutants resulting in surface concentrations orders of magnitude
above ambient levels (Lee, Shim, and Kwon 2014). Once ingested acidic gut
conditions in the gut facilitate the release and potential uptake of pollutants by
organisms (Bakir, Rowland, and Thompson 2014).
Scope of this study
The focus of this study is on two specific fish species: cod and herring, which
both have a widespread distribution in Danish waters. The goal was to analyse
the stomach contents of 100 fish from each species caught in coastal and offshore
waters of the North Sea and the Baltic Sea. It was expected that there might be a
clear differences between the exposure of coastal and offshore fish to plastic
with the latter category being less exposed. Coastal fish on the contrary would be
expected to be more exposed. The focus of this survey is on particles > 100 µm in
size and using the sampling already planned as part of DTU Aqua fish monitoring
activities.
2
Atlantic cod (Gadus morhua)
The cod is an important commercial species that is caught by commercial and
recreational fishermen throughout all Danish seas (Figure 1 and 2). Cod can
tolerate low salinities and can be found far into the Baltic Sea.
Figure 1: Distribution and density of cod > 40 cm (left) and < 40 cm (right) in the North
Sea, Skagerrak and Kattegat. Data is from 2012 IBTS monitoring cruises. The black line
marks Danish Exclusive Economic Zone.
Figure 2: Distribution and density of cod > 40 cm (left) and < 40 cm (right) in the Baltic
Sea. Data is from 2012 BITS monitoring cruises. The black line marks Danish Exclusive
Economic Zone.
Figure 3: Average
stomach contents as
percentage weight by
size class. Source: Daan,
N. (ed). 1989. Data base
report of the stomach
sampling project 1981.
Cooperative Research
Report 164. 144 pp
3
The species can grow to a maximum size of 150 cm or 40 kg, although most are
much smaller. Cod can be found at depths ranging from very shallow coastal
areas and down to 600 m. Cod are considered benthopelagic demersal fish, i.e.
living and feeding near the bottom as well as in midwaters. They feed on both
benthic as well as pelagic organisms (Fig. 3). The distribution of adult cod varies
greatly, mainly depending on the age of the individual, seasonal changes in
temperature and the distribution of prey species. At the age of 2-3 years they
become sexually mature adults. At this stage they usually remain near the sea
floor, inhabiting many different habitat types. Young cod spend most of spring
and autumn in relatively shallow water, but move to deeper waters during warm
summer months and cold winter months. As cod grow older, they generally begin
to inhabit deeper waters. Tagging experiments indicate that cod are usually quite
stationary during feeding periods, i.e. moving less than a few nautical miles per
day. During larger, one-directional migrations there are indications that cod
move at a maximum distance of up to 15 nautical miles per day (personal
communication S. Neuenfeldt, DTU Aqua, 2016).
The diet of juvenile cod is dominated by crustaceans e.g. shrimp, crabs. Larger
cod feed mainly on fish such as sandeel, flatfish, clupeids such as herring and
even juvenile cod. However, cod feed from both the sea floor and the water
column throughout their adult lives. As cod grow older, the size of preferred prey
increases (Figure 3). There are many anecdotal examples of fishermen finding
large marine litter items when gutting cod. Cod stomach retention time with a
meal consisting of sprat in the Baltic ranges between 48 to 72 hours, depending
on meal size (Andersen & Beyer 2005a; Andersen & Beyer 2005b).
Herring (Clupea harengus)
Herring is a commercially important clupeid species that is in fact made up of
many different races, which are segregated by morphology, differences in
spawning seasons, growth among other factors. They can grow to a maximum
size of 40 cm at an age of 20-25 years. Herring are schooling fish that are
completely pelagic, i.e. inhabiting and feeding only in the water column.
However, herring are demersal spawners, i.e. attaching their eggs to gravelly
substrates on the sea floor and in some cases vegetation.
Herring feed in the water column predominantly on zooplankton, which the
herring schools follow during diurnal vertical migrations. As a result, herring can
usually be found higher in the water column during the night and in deeper
waters during the day. Herring are able to use their gills to filter-feed. Herring
can also visually detect prey, such as an individual copepod or a mysid shrimp,
and attack these targets actively.
Stomach retention time for herring is markedly shorter than for cod. Almost
independently of the model used to estimate evacuation rates, stomachs can be
considered emptied after 24 hours (Darbyson et al. 2003; Bernreuther et al.
2008). No numbers exist for plastic retention times in the fish stomachs.
However, it is likely that plastic particles are evacuated from the stomach
together with other undigested remains.
4
The distribution of herring is affected by temperature, depth, frontal systems and
mixing of the water column, as well as the abundance and distribution of prey
species. Herring are present in all of the seas surrounding Denmark (Figure 4)
and different stocks are distributed throughout the entire Baltic Sea (Figure 5).
There are limited data describing the displacement rates of clupeids such as
herring and sprat. However, it has been observed in the INSPIRE project
(www.bonus-inspire.org) that Baltic Sea fishermen follow moving herring and
sprat schools for up to 15 nautical miles per day, which can be seen as an upper
limit. However, there are indications that herring are rather stationary during
feeding periods, i.e. with movement limited to approx. 2 nautical miles per day
(personal communication, S. Neuenfeldt, DTU Aqua, 2016)
Figure 4 (a): Distribution and density of herring < 20 cm (left) and > 20 cm in the
Danish seas. 2012 data from IBTS & BITS monitoring surveys.
(b):
Spatial distribution of
herring in the Baltic Sea in Quarter 4 2012 (BIAS survey). Three different stocks are
represented: Western Baltic (SDs 22-24), Central Baltic (SDs 25-29, 32) and Bothnian
Sea (SD 30). (Casini and Neuenfeldt et al., 2013)
(a)
(b)
5
Methods
Sample collection
The original sampling plan agreed upon including a collection of cod and herring
from planned DTU Aqua cruises: the International Bottom Trawl Survey (IBTS)
in the North Sea, July and August; and the Bio-C3 research cruise in the Baltic
Sea, September. The North Sea sampling in particular was plagued by poor
catches and inappropriate predetermined trawl locations. The IBTS program that
Denmark is also assigned to sample near the south-eastern North Sea coastline is
far from Danish waters. Therefore additional samples were arranged through the
Swedish IBTS monitoring cruise (Skagerrak, August) and two additional smaller
surveys; HG20 (Skagerrak, October) and TNG/SUR (Kattegat, November).
Catches of North Sea herring were supplemented by catches from commercial
vessels (H218-H10, S349, RI366) and catches from another cruise (SOLEA). An
overview of acquired and analysed fish is given in Table 1.
Table 1: Summary of the number of fish stomachs analysed by the project split
between regions. Coastal is defined as being within 12 nm off the coast. Numbers
in brackets indicate number of fish obtained.
Cod
Region
Cruise / Vessel
Coastal
Offshore
Total
North Sea incl.
Kattegat/Skagerak
IBTS-DK, IBTS-SE,
TNG/SUR, HG20
28 (28)
72 (143)
100 (171)
Baltic
BIO-C3
51 (53)
50 (97)
101 (150)
Total
79 (81)
122 (240)
201 (321)
Herring
Region
Cruise
Coastal
Offshore
Total
North Sea incl.
Kattegat/Skagerak
IBTS-DK, IBTS-SE,
Solea, H218-H10,
S349, RI366
50 (63)
50 (111)
100 (174)
Baltic
BIO-C3
55 (78)
50 (95)
105 (173)
Total
105 (141)
100 (206)
205 (347)
Figures 5 - 8 show location of Cod and Herring caught in the North Sea and Baltic,
respectively, that were later analysed for microplastic. The coloured dots
indicate stations with number of fish analysed (if no number exists: station was
not chosen for analysis). Light blue areas represent 12 nautical miles zones, the
orange area the Danish exclusive economic zone (EEZ) and the orange grid ICES
statistical rectangles with names. Maps were produced with QGIS
(http://qgis.org).
6
Figure 5:
Location of cod catches in the North Sea.
Figure 6:
Location of cod catches during the Baltic Sea sampling.
7
Figure 7:
Location of herring catches in the North Sea.
Figure 8: Location of herring catch in the Baltic Sea
Sample processing
For all cruises except the BIO-C3 and IBTS cruise, the fish were frozen
immediately after catching and the stomachs were extracted later on return to
the laboratory on land. During the BIO-C3 cruise the extraction was partly
carried out in the laboratories on board the ship. Cod stomachs were extracted
and transferred to zipper bags under clean conditions ensuring minimal
exposure time and potential contamination.
Sample digestion
Initially the project followed the recently published guidelines for isolating
particle in fish stomachs provided by ICES (ICES 2015) as recommended by the
Nature Agency in the contract. However the digestion mixture recommended
8
was found to be much too harsh and readily dissolved a wide range of polymer
types. A study documenting this was therefore carried out and is provided in the
appendix. An alternative approach based on what was used in the finally
developed digestion method was a result of test series to optimise tissue
digestion and removal of fat and oil residues from the samples, as well as the
protection of contained microplastics of all major commodity plastics. The
protocol was inspired by experiences from an earlier Nature Agency project
(Sørensen et al. 2013) and similar recent work (Agersnap 2013; Strand et al.,
unpublished).The digestion of the stomach tissue was carried out in acid
washed glass jars. A digestion solution of 150 ml KOH (1120 g/L) and 150 ml
NaClO (14% active chlorine) to 700 ml water was prepared and filtered through
a 30 µm filter.
For each stomach sample 5 ml stock solution per 1 g stomach wet weight was
dispensed and the lid closed loosely (i.e. not gas tight). The jars were
subsequently treated with 10 minutes ultrasound bath and 1 hour on a shaker
table. If the stomach tissue was still visible the period on the shaker was
extended. Microplastic particles were isolated from the digestion fluid by
vacuum filtering through a metallic sieve stack consisting of a mesh of 1 mm and
300 µm and a 100 µm polyamid filter (plankton net) and rinsed with MilliQ
water.
Microplastic particles retained on each of the mesh size were classified with
respect to shape, colour and size using light microscopy. All microplastic samples
were tested using a melting device to confirm their plastic origin. Additionally a
sub-fraction from each category was characterised using Raman spectroscopy to
allow polymer identification (Lenz et al. 2015).
Results and Discussion
Distribution of plastic between areas and species
Of the 72 offshore North Sea cod analysed 49% were found to have microplastic
in their stomachs. Whereas from the 28 coastal North Sea cod 14% contained
microplastic. The respective numbers for the Baltic cod were 26% and 16%
(Table 2). There was an overall tendency for higher likelihood of stomach plastic
content for the offshore cod.
A similar pattern was apparent for the Baltic herring. For the 55 herring sampled
from coastal Baltic waters, plastic was found in only 4 fish (7%) (Table 3). For
offshore herring from the Baltic twice as many had plastic (16%). For herring
from the North Sea the number of fish with microplastic in their stomach content
was notably higher, 30 and 16% for coastal and offshore respectively.
The findings presented do not support the hypothesis that fish from coastal
stations are more exposed to land-based sources of microplastic and would
therefore be contaminated to a higher degree. The opposite is found in the
present data for 3 out of 4 groups (Figure 9). There was no significant correlation
9
between distance from shore and microplastic content in fish from all stations
analysed.
North Sea herring was the only group to have more coastal fish containing
microplastic. It must be noted that stations marked as coastal and offshore were
in fact often in close proximity to one another – a factor which makes robust
comparisons between coastal and offshore difficult.
Table 2. Summary of the results for the analyses on cod stomachs
Overall microplastic containing fish among cod amounts to 39% in the North Sea
and 21% in the Baltic, for herring to 23% and 11%, respectively. There was a
general tendency for higher numbers of plastic in larger stomachs (i.e. the large
cod stomachs) which seems reasonable simply because more material is
available for analysis. However, evaluating the microplastic load relative to the
weight of the fish stomach, herring shows a roughly four times higher abundance
(Figure 10). Comparing fish body length to microplastic load showed no clear
correlation, despite the same slight tendency that highest number of plastic
10
items were generally observed in fish in the upper half of the group's size
spectrum (Figure 11).
In the Baltic Sea, fish were analysed from the Bornholm Basin and the Eastern
Gotland Basin / Bay of Gdansk. Figure 12 is providing an overview of the
percentages of fish that were found containing microplastic in their stomach in
the two respective regions. For cod and herring from the Bornholm basin,
relatively more fish were contaminated with plastic. However, it must be noted
that the study was intended to focus on Danish waters as much as possible,
resulting in small sample sizes from the Gotland basin (n= 24 for cod and n= 10
for herring).
Table 3. Summary of the results for the analyses on herring stomachs
11
Figure 9. :
Summarising bar diagram illustrating microplastic occurrences in Cod and
Herring form the North Sea and Baltic, respectively. Sample sizes can be taken from
Table 2 and 3 (analysed total)
Figure 10:
Box plots showing
the relative microplastic load
per gram fish stomach.
Depicted are the median values
with 25 and 75% percentiles.
Error bars indicate maximum
and minimum of the data set.
12
Figure 11:
Relation between size of fish (total length) and microplastic particles and
fibres. Blue for coastal, red for offshore stations
Figure 12:
Comparison of percent fish containing one or more pieces of microplastic
from Bornholm area and south-eastern Baltic. Cut-off part of the pie charts represents
plastic-containing fish.
13
Plastic categories
Some examples of the types of plastic isolated from fish stomachs are shown in
Figure 13. From the 95 pieces of plastic discovered the vast majority (83%) have
been fibres ranging in length from 0.15 to 57 mm. Particle sizes have ranged
from 0.1 to 5.6mm. The largest piece was found in a cod from the Skagerrak
(Figure 13, right).
Figure 13: Examples of the different types of objects found in the stomachs of the fish
examined. To the left a typical fragment/particle, in the middle a fibre and on the right
an extreme example of an approximately 5 cm long piece of rubber or silicon originating
from a sport fishing bait found in the stomach of a cod
The results of the colour analysis are shown in Figure 14a showing a similar
colour distribution across herring and cod whether them being from the Baltic or
the North Sea. Microplastic colour differences are observed between coastal and
offshore fish (Figure 14b). Both herring and cod from rather coastal zones
appear to contain microplastic of less colour variability than its offshore
equivalent.
Figure 14: Microplastic colour variability. Whole bars represent 100% of collected
micrplastic in each group. Colours were recorded as grey, green, blue,
transparent/whitish, yellow, red/pink, black. (a)
Colour distribution in cod C and
herring H from North Sea and Baltic.
(b)
Colour distribution split up for off-shore and
coastal samples.
(a)
(b)
14
Raman Analysis
The polymer type of a small subsample was analysed in a Raman micro-
spectrometer. Exemplary spectra are shown below (Figure 15). Some common
polymer references could be matched to a high degree, however in other cases
no analysable spectra could be obtained from particles or fibres that otherwise
exhibited clearly plastic-like features (morphology, melting behaviour).
A given polymer type found in an ingested microplastic particle cannot be used
to trace litter sources. This is due to the fact that the majority of plastic types is
applied across all continents and sectors such as the building industry, consumer
products, packaging, etc. Shapes such as fibrous microplastics, usually made of
polyester, polyamide or of acrylic nature, originate mostly from textiles.
However there are no numbers on how much is lost during production compared
to consumer waste (from washing machines).
The labour intensive nature of measuring individual particles limits the
feasibility of quantitative spectroscopic analysis. Raman signal confounding
factors such as fluorescence overlay from strongly Raman-active additive
compounds e.g. colourants, degradation state or surface coatings of microplastics
are problems that require further development and adaptation of Raman
techniques to be more efficiently harnessed for marine microplastic detection
and characterisation. The methods and labour time available for this study did
not allow for a quantitative microplastic identification based on spectrometric
analysis. The presented results are only covering a few selected particles that
were in a good measurable condition. The results demonstrate the current
limitations but at the same time the fundamental possibilities of Raman
microspectroscopy assisted microplastic analysis. Currently, we are intensifying
our research in automated spectroscopic microplastic identification where
Raman and other techniques will be developed further for rapid microplastic
sample analysis, which is urgently needed in all studies investigating
microplastic pollution from environmental samples.
FTIR microscopy, an alternatively used spectral identification technique, would
require the same if not higher degree of sample purification. It will not be
confounded by fluorescent additives but is inferior in identification of dark, light
absorbing materials. Both techniques together can give complementary results
and are commonly used in combination in analytical chemistry. However,
regardless of which spectroscopic technique is applied, there is a fundamental
need to develop or adapt automatised procedures for microplastic sample
analysis before larger studies on chemical composition of marine microplastic
litter becomes cost-efficient.
The proportion of fibre to particles was slightly higher in coastal regions
compared to offshore. This appears reasonable as microplastic fibres usually
arise from fabric fragments which pass washing machines and waste water
treatment plants (Browne et al. 2011). The mentioned difference between the
ratios (fibre:particle) 3.7 and 2.9. It has to be noted that most stations that are
within the 12 nautical mile zone and declared coastal are still relatively far out at
sea which was due to the rather rigid planning of the monitoring cruises.
15
The overall ratio of fibres to particles was markedly higher in herring compared
to cod, 6.4 and 2.4, respectively. This could reflect their feeding strategy where
when filter feeding fibres are held back by the gills. A previous study conducted
by DTU Aqua in 2013 which analysed 90 whiting and herring from the Belt Sea
for microplastic (>0.5 mm) found 31% and 27% in the digestive systems. Also
here fibres were the prevailing type of plastic ranging from 0.5 to 4 mm
(Sørensen et al. 2013).
Figure 15:
Raman spectra of four plastic items found in fish stomachs (microscopic
photographs on the right, scale bars in micrometre). The spectra are each shown with a
reference spectrum obtained from commercially available consumer products of known
polymer type. From top to bottom: polyurethane, polystyrene, polypropylene and
polyethylene.
16
A similar pattern can be seen in this study where cod reached an overall
microplastic occurrence of 30% and herring 17%. Thus it is likely that
microplastic can be found in the stomachs of up to about a quarter of fish in
Danish waters. The implications of this are currently unknown and the focus of
much research and discussion.
With respect to monitoring of microplastic pollution it would be relevant to
include a study on fish from very coastal habitats. This could possibly be done
near primary microplastic input areas such as large cities and industrial sites or
secondary microplastic accumulation zones mainly influenced by present
hydrodynamics and density of the specific plastic. Accumulation zones for the
North Sea and Baltic region are not fully documented yet and the mapping of
macro and microplastic there should be subject to coming research projects in
order to identify hotspot areas.
Development of an indicator
The widespread distribution and differing feeding strategies of both species
make them ideal for development into indicator species for microplastic
ingestion by fish. Additionally the fact that they are both part of routine
monitoring surveys and intensively studied provides added economical and
scientifically benefit. Comparing the two species, herring is by far the easiest to
process due to the smaller stomach size and shorter base-digestion time in the
laboratory.
The results also showed that herring has a higher load of microplastic per
stomach weight compared to cod (Figure 10). This could be due to the feeding
strategy of herring which may bias microplastic uptake. In areas with high prey
abundance herring often switches to filter-feeding mode which enhances
chances for accidental uptake as shown for mussels and zooplankton. It is also
possible that herring mistake microplastic for prey. However, the size of particles
found was usually in the sub-millimetre range, far below the usual feeding size of
herring. This suggests that plastic entered the fishes accidentally through
filtering or along with ingested prey.
When developing an indicator it is very important to clearly specify the goal. In
Danish waters to date herring could be used to give indication for trends in
microplastic load in pelagic fish species. It would however be beneficial to have a
reference beyond local anthropogenic (e.g. oceanic) influence to compare to and
distinguish between local influence and general widespread pollution. In
addition, before this can reach management levels there needs to be more
knowledge of the seasonal variation within microplastic concentrations. For one
reason, the fish migrates to coastal zones during spawning seasons which could
change the uptake tremendously. Secondly, the pelagic microplastic load in the
Baltic Sea is likely to vary with freshwater input and mixing which changes
markedly seasonally. Based on the described migration capabilities and stomach
retention times it can be expected that microplastics from herring stomachs
originate from sites within a maximum of 15 nautical miles from the fish
sampling location. For cod the range can be expected to be similar. It should be
17
noted that no direct data on stomach retention times of microplastics have been
measured for the two species. Although plastic is expected to generally pass the
intestines with any other non-digestible matter it is possible that special shape
and morphological properties of microplastics can lead to longer retention.
Fibres or highly edged particles might get stuck in grooves and narrow sections
inside the intestines, and worst case entangle permanently.
A current study at DTU Aqua is analysing herring and sprat for microplastic and
simultaneously taken water samples tracing back from 2015 to the late 1980's,
including a comparison between seasons. While there has been research efforts
on determining the plastic ingestion by planktivorous fish in several waters
worldwide, to our knowledge there are no investigations focusing on the
abundance, distribution and composition of plastic litter over a longer period of
time and whether the stomach content reflects the plastic concentrations in the
water. This could give valuable insights into correlations between seasons,
changes in microplastic density and sprat as a possible indicator species and thus
a step forward towards developing an indicator species.
Conclusions
The results of this report present the microplastic (>100 µm) load of cod and
herring from the North and Baltic Sea which on average amounts to 23% of all
fish. Contrary to expectation we have not seen differences between offshore and
coastal caught fish. For future investigations one could consider the analyses of a
very coastal population either herring or short spined sea scorpion (dk.: ulk, e.g.
not more than 1 nautical mile from shore). Part of this could be to compare
between heavily contaminated areas such as harbours, big cities, industries, and
sparsely populated regions. Cod as well as herring are appropriate indicator
species for microplastic pollution in demersal/benthopelagic (i.e. living near the
seafloor as well as in the water column) and pelagic fish species, respectively.
However due to stomach size, cod stomach tissue digestion is a more laborious
procedure. For future work inclusion of a reference population from an oceanic
site such as the open North Atlantic, would facilitate the distinguishing between
local and more diffuse global pollution.
Acknowledgements
This study was funded by the Ministry of Environment and Food of Denmark’s
Nature Agency (Naturstyrelsen). The BIO-C3 cruise on RV DANA in September
2015 was funded by the Danish Centre for Marine Research (grant 2015-04). The
authors would like to thank Bastian Huwer for the logistic support and
organisation of the cruise as well as Zac Calef for collecting fish for the analysis.
Further, the authors would like to thank Frank Ivan Hansen, Kai Wieland,
Susanne Hansen, Barbara Bland, Peter Vingaard Larsen, Dennis Ulrik Andersen,
Thomas Møller, Jan Pedersen, Helle Rasmussen, Aage Thaarup and Stina Bjørk
Stenersen Hansen for organising and collecting fish from supplementary cruises.
18
References
Agersnap, Sune. 2013. “Mikroplastik I Havmilløet.”
Andersen, N. G., & Beyer, J. E. (2005a). Gastric evacuation of mixed stomach
contents in predatory gadoids: an expanded application of the square root
model to estimate food rations. Journal of Fish Biology, 67(5), 1413-1433.
Andersen, N. G., & Beyer, J. E. (2005b). Mechanistic modelling of gastric
evacuation in predatory gadoids applying the square root model to describe
surface‐dependent evacuation. Journal of fish biology, 67(5), 1392-1412.
Anon. 2008. “Directive 2008/56/EC of the European Parliament and of the
Council of 17 June 2008 Establishing a Framework for Community Action in
the Field of Marine Environmental Policy (Marine Strategy Framework
Directive).” Offical Journal of the European Union 164: 19–40.
Bakir, Adil, Steven J Rowland, and Richard C Thompson. 2014. “Enhanced
Desorption of Persistent Organic Pollutants from Microplastics under
Simulated Physiological Conditions.” Environmental Pollution (Barking,
Essex : 1987) 185 (March). Elsevier Ltd: 16–23.
doi:10.1016/j.envpol.2013.10.007.
Barnes, David K a, Francois Galgani, Richard C Thompson, and Morton Barlaz.
2009. “Accumulation and Fragmentation of Plastic Debris in Global
Environments.” Philosophical Transactions of the Royal Society of London.
Series B, Biological Sciences 364 (1526): 1985–98.
doi:10.1098/rstb.2008.0205.
Bernreuther, M., Herrmann, J. P., & Temming, A. (2008). Laboratory experiments
on the gastric evacuation of juvenile herring (Clupea harengus L.). Journal of
Experimental Marine Biology and Ecology, 363(1), 1-11.
Browne, M.A., Crump, P., Niven, S.J., Teuten, E., Tonkin, A., Galloway, T. and
Thompson, R. (2011). Accumulation of Microplastic on Shorelines
Woldwide: Sources and Sinks. Environ. Sci. Technol., 2011, 45 (21), pp 9175–
9179
Casini and Neuenfeldt et al., Study Group on Spatial Analyses for the Baltic Sea
(SGSPATIAL), ICES report, 2013.
Cole, Matthew, Pennie Lindeque, Elaine Fileman, Claudia Halsband, Rhys
Goodhead, Julian Moger, and Tamara S Galloway. 2013. “Microplastic
Ingestion by Zooplankton.” Environmental Science & Technology 47 (12):
6646–55. doi:10.1021/es400663f.
Daan, N. (ed). 1989. Data base report of the stomach sampling project 1981.
Cooperative Research Report 164. 144 pp
Darbyson, E., Swain, D. P., Chabot, D., & Castonguay, M. (2003). Diel variation in
feeding rate and prey composition of herring and mackerel in the southern
19
Gulf of St Lawrence. Journal of Fish Biology, 63(5), 1235-1257.
Fendall, Lisa S, and Mary a Sewell. 2009. “Contributing to Marine Pollution by
Washing Your Face: Microplastics in Facial Cleansers.” Marine Pollution
Bulletin 58 (8). Elsevier Ltd: 1225–28.
doi:10.1016/j.marpolbul.2009.04.025.
Galgani, F, and G Hanke. 2013. “Marine Litter within the European Marine
Strategy Framework Directive.” ICES Journal of Marine Science 70: 1055–64.
doi:10.1093 / icesjms / fst122.
Gregory, Murray R. 1996. “Plastic ‘scrubbers’ in Hand Cleansers: A Further (and
Minor) Source for Marine Pollution Identified.” Marine Pollution Bulletin 32
(12): 867–71. doi:10.1016/S0025-326X(96)00047-1.
Ivar do Sul, Juliana a, and Monica F Costa. 2014. “The Present and Future of
Microplastic Pollution in the Marine Environment.” Environmental Pollution
(Barking, Essex : 1987) 185 (February). Elsevier Ltd: 352–64.
doi:10.1016/j.envpol.2013.10.036.
Lee, Hwang, Won Joon Shim, and Jung-Hwan Kwon. 2014. “Sorption Capacity of
Plastic Debris for Hydrophobic Organic Chemicals.” The Science of the Total
Environment 470-471 (February). Elsevier B.V.: 1545–52.
doi:10.1016/j.scitotenv.2013.08.023.
Lenz, Robin, Kristina Enders, Colin A Stedmon, David M A Mackenzie, and Torkel
Gissel. 2015. “A Critical Assessment of Visual Identification of Marine
Microplastic Using Raman Spectroscopy for Analysis Improvement.” Marine
Pollution Bulletin. Elsevier Ltd. doi:10.1016/j.marpolbul.2015.09.026.
NST. 2012. Danmarks Havstrategi - Basisanalyse. Danmarks Havstrategi -
Miljømålsrapport. 100 pp.
NST. 2014. Danmarks Havstrategi Overvågningsprogram.
Setälä, Outi, Vivi Fleming-Lehtinen, and Maiju Lehtiniemi. 2014. “Ingestion and
Transfer of Microplastics in the Planktonic Food Web.” Environmental
Pollution (Barking, Essex : 1987) 185 (February). Elsevier Ltd: 77–83.
doi:10.1016/j.envpol.2013.10.013.
Sørensen, Thomas Kirk, Colin Stedmon, Kristina Enders, and Ole Henriksen.
2013. “Analyse Af Marint Affald I Sild Og Hvilling Fra Det Nordlige
Storebælt,” no. December.
Strand, J., P. Lassen, Y. Shashoua, and J.H. Andersen. “Occurrence of Microplastic
Particles and Bio-Geochemical Relationships in Sediments from Danish
Waters (in Prep.).”
TSG-ML. 2014. “Executive Summary 2013 Directive Executive Summary of the
Guidance on Monitoring of Marine Litter in European Seas,” no. January
2016.
20
Wright, Stephanie L, Richard C Thompson, and Tamara S Galloway. 2013. “The
Physical Impacts of Microplastics on Marine Organisms: A Review.”
Environmental Pollution (Barking, Essex : 1987) 178 (July): 483–92.
doi:10.1016/j.envpol.2013.02.031.
21
Appendix 1
Assessment of existing digestion approaches for
isolation of marine microplastic from biota
Introduction
Plastic is accumulating in the oceans where it, exposed to various environmental
stressor (UV, waves, sand, salt), breaks down to ever smaller fragments such as
microplastic. Every size of plastic is of potential harm to a corresponding feeding
size spectrum of a particular group of marine species. The quantification of
microplastic in stomachs of various biota is central to assessing the impact and
extant of plastic pollution in the environment to organisms and to determining
its pathways through food webs and sinks. The development of techniques to
isolate and characterise microplastic is, however, still under in progress and
methodologies are across different studies are barely standardised. The
monitoring of plastics in fish has been integrated into the European Union
Marine Strategy Framework Directive (MSFD) and Oslo-Paris Commission
(OSPAR) guidelines.
Before microplastic distribution and impacts can be systematically assessed
standard sampling and laboratory techniques are required. A major challenge is
the removal of the organic matter that is often associated with microplastic
particles. This can be both detrital, for environmental samples as the
hydrophobic nature of plastics will cause aggregation, but also organic matter in
the form of tissue in the case of samples from biota. Before microplastic can be
characterised either visually or spectroscopically the organic matrix must be
removed.
Various approaches have been made using enzymes, a range of different bases
(KOH, alkaline cleaning agents) and acids (HNO3 HClO4 ), oxidizer (H2O2).
A preliminary protocol for the digestion of organic matter in conjunction with
microplastic isolation and sample preparation has been provided issued by the
International Council for the Exploration of the Seas (ICES). The procedure
recommends the usage of a mixture of HNO3:HClO4 (4:1) as digestive agents [1].
During preparations for national monitoring activities for microplastics in fish
stomachs we found indications that the recommended acid combinations had
severe effects on a range of common polymers. Here we report the results of a
set of 20 different common polymers that as a result of the tissue digestion
treatment during the recommended exposure time some polymers either
completely dissolved, others partly disintegrated, or changed colour (surface
damages) or were resistant.
Various studies have already applied the recommended acid mixture [2]–[4] or
nitric acid alone [5] and should therefore be handled with caution. A validation of
22
different applied chemicals for method improvement is highly needed and was
accomplished in this study by exposing the same test polymers to KOH [6] in
combination with NaClO (14% free Chlorine) [7] as well as VIP1 (with 30%
NaClO) [8], [9].
Methods
Plastic resistance to test chemicals
Polymer samples were chosen from a selection of consumer plastic items of
which the material or recycling label was visible. Several pieces of about 0.5 cm
size were cut off each item and transferred into a separate 8 ml laboratory glass
vial with a black butyl/PTFE screw top lid. Treatment solutions (5 ml) were
added and vials were kept upright during the testing period. Four different
digestion solutions were compared (Table 1). The effects of the different
treatments were documented photographically before the addition of chemicals
and after 30 min, 1 h and 5 h. Following this treatment period at room
temperature all samples were exposed to 80°C for 20 min and observed changes
thereafter documented in the same manner. In the case of the acid mix treatment
an extra image series was recorded after 10 h to better document also weak
changes. Observed effects were categorised by severity into four levels (Table 2).
Further notes were taken to better describe particular effects observed.
Table 1: Description of the four digestions solutions tested.
Acid mix
4:1 (v:v) HNO3 69% (AnalR, VWR International S.A.S.) + HCLO4 70%
(Rectapur, VWR International S.A.S.), procedure: Addition of 5 ml/g and
digestion for 5 hours and subsequent heating for 10 min in 80°C.
KOH
Solution of KOH pellets (Emsure, Merck) in microfiltrated H2O 1120 g/L,
procedure: see above.
NaClO
NaClO solution, 14% active chlor (VWR International S.A.S.), procedure: see
above.
Industrial
CIP agent
VIP 1 (Novadan Aps), ready solution containing ~3% Potassium hydroxide,
~1% Potassium tripolyphosphate, ~1% Potassium silicate and ~7% Sodium
hypochlorite, procedure: see above.
Table 2: Definition of the four different impact levels observed.
Level of impact
Description
L1
Beginning visual recognisable changes (colour, surface morphology)
L2
Morphological changes, beginning dissolution
L3
Strong morphological disintegration, change of bulk structure
L4
Complete dissolution / disintegration
23
Raman micro-spectrometry
In order to better evaluate weak changes on the outer matrix of the polymer a
range of polymers was measured with Raman micro-spectrometry after the
digestive treatments. The spectra were compared to a spectra library of the same
polymers in untreated condition. Table 4 shows the respective polymer /
digestant combinations tested.
Testing tissue digestion effectiveness
A comparison study was conducted in order to test for digestion effectiveness
among the KOH (I), NaClO (II), KOH and NaClO in combination (III) and VIP1 (IV).
VIP1 is a ready-made solution which contains both potassium hydroxide (KOH,
3%) and sodium hypochlorite (NaClO, 7%) as the main ingredients which is why
these chemicals were tested for their digestive power first separately and later in
combination. The test stomach tissues weighed between 8 and 22gram. Per gram
of tissue, 5ml test solution was added to the sample. All treatments were first
subjected to a 15 minutes ultrasonic bath followed by two hours of thorough
shaking. The most effective digestion solution was then further diluted to find an
optimum i.e. most economical concentration.
Results
Acid mix treatment
Exemplary observations from the acid mix exposure are shown in Figure 1 – 4
and summarised in Table 3. The strongest effects were observed for Polyamid
(PA), Polyurethan (PU) and black tire rubber elastomer, all of which were
completely dissolved by the acidic treatment. In case of PA6 the complete
dissolution (L4) was observed within seconds to minutes after submersion in the
acid mix. Other structurally affected polymers were Acrylonitrile butadiene
styrene (ABS), Poly(methyl methacrylate) (PMMA) and Polyvinyl chloride (PVC).
The latter one being affected in moderation, mainly colour leaching and softening
(L2 – L3). Polymer samples of Polycarbonate (PC), expanded and solid
Polystyrene (PS) and Polyethylene terephthalate (PET) were structurally little
affected. Only staining or colour loss could be observed visually (L1). No effects
were observed for Polypropylene (PP), high density and low density
Polyethylene (HDPE, LDPE), Ethylene-vinyl acetate (EVA) and
Polytetrafluoroethylene (PTFE). The heating to 80°C after the digestion period
was fund to have an exacerbating effect on the polymer destruction in all cases
where an effect has been observed beforehand.
Raman micro-spectrometry that has been performed on polymers which did not
show severe visible changes after the acid mixture treatment revealed that apart
from ABS all remaining polymers gave recognisable spectra although some
showed signs of degradation or peak shifts (Table 4).
24
Figure 1:
ABS bloating by acidic treatment after 5 h (left), 10 h (middle) and additional
heating to 80°C (right).
Table 3: Tested polymer types in the acid test treatment with observed impact
levels.
0.5 h
1 h
5 h
10 h
80°C
no change
PP
x
LDPE
x
HDPE
x
PS
L1
EPS
L1
ABS
L1
L2
L3
PU
L2
L3
L4
PA
L4
PA
L2
L3
EVA
x
PET
L1
PC
L1
Nitrile
L1
L3
L4
PVC 1
L1
L2
PVC 2
L1
L2
PVC 3
x
PMMA
L2
L3
PTFE
x
RB
L3
L4
RB
L2
L3
L4
TR
L1
L2
L3
L4
25
Alkaline Treatments
All tested polymers did not show any impact according the defined levels 1 to 4
during the treatments using KOH, NaClO, VIP1 alone or in the described
combinations. The characteristics of acquired Raman spectra were not or little
changed after the polymer samples were treated with the 30% dilution of the 1:1
ratio mixed KOH:NaClO solution. For the pure and saturated KOH solution strong
spectral deviations and lower quality spectra were obtained, however, all
Figure 2: PU before (left) and after 5 h of acidic treatment.
Figure 3: PA-6 before (left) and after first contact with acidic
treatment.
Figure 4: No observable changes after acidic treatment of
polyolefins (PP left, HDPE right).
26
polymer types could be recognised by means of comparing against the library of
spectra from the untreated polymers.
Table 4 Evaluation of spectra changes after chemical treatments via Raman.
Legend: Recognisable, widely identical; Recognisable, with noted peak changes
or flourescense; Hardly recognisable; Not recognisable; n/a: not measured.
Acid Mixture
KOH
30 % KOH : NaClO
PP
Recognisable, widely
identical
Recognisable, widely
identical
Recognisable, widely
identical
LDPE
Recognisable, alterations
from the reference are
likely to not be from the
acid but due to variances
among LDPE.
Recognisable
Recognisable, widely
identical
HDPE
Recognisable, widely
identical
Recognisable, widely
identical
Recognisable, widely
identical
PS
Recognisable, slightly
noisy, florescence
Recognisable, widely
identical
Recognisable, widely
identical
EPS
n/a
n/a
Recognisable, widely
identical
ABS
Not recognisable, change
in peak positions and
florescence. 1605 peak
remains, 1353 peak
occurred
Recognisable, slightly
noisy
Recognisable
PA
n/a
Recognisable, missing
peak at 147
Recognisable, peak at 147
weakened
PET
Recognisable, widely
identical
Recognisable, peak at
3080 and 861 enhanced,
new peaks at 1418 and
1132
Recognisable, widely
identical
PC
Recognisable, widely
identical
Recognisable, new peak
at 1065
Recognisable, widely
identical
PVC
Recognisable,
Fluorescence, weaker
main peaks at 704 and
637 indicating
degradation processes
Recognisable, weaker
main peaks at 704 and
637 indicating
degradation processes
Recognisable, widely
identical
PMMA
Recognisable, new peak
at 1054 and 1308
Recognisable, new peak
at 1070
Recognisable, widely
identical
PTFE
n/a
n/a
Recognisable, widely
identical
27
Testing tissue digestion effectiveness
Table 5 shows that among all chemicals tested KOH in combination with NaClO
had the most satisfying effect i. e. it dissolved the sample tissue completely. KOH
or NaClO alone are not nearly as effective. VIP1 was closest to the result of KOH:
NaClO, however some slimy sediment remained. KOH:NaClO was then tested in
three different (IIIa, IIIb, IIIc) proportions of which IIIc (1:1) was found to be
optimal. Higher proportions of NaClO causes foam formation, too little reduces
the digestion effectiveness (Table 6).Eventually a dilution of the KOH: NaClO (1:1)
mixture to 30% was tested due to economical reasons and found to still enable a
full digestion. While preparing this solution water should be added first before
the two reagents are added to avoid precipitation.
Table 5 Descriptive test results for digestion effectiveness of KOH (I), NaClO (II),
KOH: NaClO (III) and VIP1 (IV). a and b indicate when different concentrations
were applied. The most effective treatment is marked in bold.
Treatment
Post treatment
I
100% KOH
layer of black/brown slime afloat, no big pieces
IIa
30% NaClO
milky, stomach still floating as one piece
IIb
100% NaClO
only half of mixture added because of very strong foam formation,
partly loss of sample.
III
1:1
(KOH:NaClO)
sample tissue completely dissolved
IVa
30% VIP1
sample mostly dissolved, slimy sediment remaining
IVb
100% VIP1
sample dissolved better, less slimy sediment remaining
Table 6 Testing different proportions of the KOH:NaClO mixture. The most
effective one is marked in bold.
Treatment
Post treatment
IIIa
2:1
(KOH:NaCLO)
many small pieces (food remaining e.g. copepods )
IIIb
1:2
(KOH:NaCLO)
Foam formation (minimized by cooling glass bottle in cold water),
few small pieces (food remaining)
IIIc
1:1
(KOH:NaClO)
Foam formation (slightly less than above). Sample tissue
completely dissolved
Conclusion and Recommendations
As we tested and evaluated the effect of the acid mixture on macroscopic plastic
items one can assume that when exposing microplastic to the chemicals the
destructive effect is even more severe due to its larger surface area and rather
fragile nature. This is especially relevant to studies investigating microplastic
below 300 µm. Study results based on the use of the ICES protocoll
recommending the mixture of HNO3 and HCLO4 should therefore interpreted
28
with precaution.
Another study (Strand, in prep.) successfully used VIP1 for sediment and tissue
digestion purposes. The VIP1 used in this study was already several month old
and has been opened before which makes it likely that active chlorine has
deteriorated to some extent. This would mean that the effectiveness was limited
and perhaps better results can be obtained when using a fresh VIP1 solution.
However, its access is more difficult for a wider scientific community (produced
in DK) and it contains ingredients of which an auxiliary effect is unknown to us.
Because of that and due to the convincing digestion effectiveness the usage of a
30% KOH:NaClO mixture (i.e. for 1 litre: 150 ml saturated KOH solution 1120 g/l
+ 150 ml NaClO solution with 14% active chlorine + 700 ml microfiltrated water)
was found most appropriate and we suggest it therefore as a fast, inexpensive
and effective digestion method. We recommend it being further compared under
aspects of economic and expedient test execution against existing protocols of
enzymatic digestion. Having evidenced the complete dissolution in reasonable
work time while not impairing the integrity of all important plastic polymer
groups, we argue for the described method being considered in international
guidelines when targeting standard protocols for worldwide usage.
References
[1] ICES, “ICES Special Request Advice Northeast Atlantic and Arctic Ocean.
OSPAR request on development of a common monitoring protocol for
plastic particles in fish stomachs and selected shellfish on the basis of
existing fish disease surveys.,” IICES Advice 2015, B. 1, no. June, pp. 1–6,
2015.
[2] Z. V. Mechtild, “Impact of microplastic uptake : contamination in sprat and
microplastic-mediated uptake of PAHs by European shore crab,” 2015.
[3] V. Griet, V. C. Lisbeth, J. C. R, M. Antonio, G. Kit, F. Gabriella, K. M. J. J, D.
Jorge, B. Karen, R. Johan, and D. Lisa, “A critical view on microplastic
quantification in aquatic organisms,” no. 2014, 2015.
[4] B. De Witte, L. Devriese, K. Bekaert, S. Hoffman, G. Vandermeersch, K.
Cooreman, and J. Robbens, “Quality assessment of the blue mussel (Mytilus
edulis): Comparison between commercial and wild types,” Mar. Pollut.
Bull., vol. 85, no. 1, pp. 146–155, 2014.
[5] L. Van Cauwenberghe and C. R. Janssen, “Microplastics in bivalves cultured
for human consumption.,” Environ. Pollut., vol. 193, pp. 65–70, 2014.
[6] E. M. Foekema, C. De Gruijter, M. T. Mergia, J. A. van Franeker, A. J. Murk,
and A. a Koelmans, “Plastic in north sea fish.,” Environ. Sci. Technol., vol. 47,
no. 15, pp. 8818–24, Aug. 2013.
[7] T. K. Sørensen, C. Stedmon, K. Enders, and O. Henriksen, “Analyse af marint
affald i sild og hvilling fra det nordlige Storebælt,” 2013.
[8] S. Agersnap, “Mikroplastik i havmilløet,” 2013.
[9] J. Strand, P. Lassen, Y. Shashoua, and J. H. Andersen, “Occurrence of
microplastic particles and bio-geochemical relationships in sediments
from Danish waters (in prep.).”
29
