Available via license: CC BY-NC-ND 4.0
Content may be subject to copyright.
A comprehensive investigation of industrial plastic pellets on beaches
across the Laurentian Great Lakes and the factors governing
their distribution
Patricia L. Corcoran
a,
⁎,Johanna de Haan Ward
b
,Ian A. Arturo
a
, Sara L. Belontz
a
, Tegan Moore
c
,
Carolyn M. Hill-Svehla
d
, Kirsty Robertson
c
, Kelly Wood
c
, Kelly Jazvac
e
a
Department of Earth Sciences, University of Western Ontario, London, ON, Canada
b
Department of Statistical and Actuarial Sciences, University of Western Ontario, London, ON, Canada
c
Department of Visual Arts, University of Western Ontario, London, ON, Canada
d
Surface Science Western, University of Western Ontario, London, ON, Canada
e
Department of Studio Arts, Concordia University, Montreal, QC, Canada
HIGHLIGHTS
•Shorelines of the GreatLakes are littered
with industrial plastic pellets.
•42 of 66 beaches contained pellets, for
an average of 19.1 pellets/m
2
.
•Abundance increased with number of
plastic industries and proximity to
river mouths.
•Variety was greatest on a beach in a wa-
tershed containing 112 plastic indus-
tries.
•Pellets were most abundant on very
fine, fine, and medium sand beaches.
GRAPHICAL ABSTRACT
abstractarticle info
Article history:
Received 11 April 2020
Received in revised form 21 July 2020
Accepted 23 July 2020
Available online 25 July 2020
Editor: Damia Barcelo
Keywords:
Plastic pellets
Plastic industry
Laurentian Great Lakes
Pellet spills
River mouth
Grain size
Industrial, pre-consumer pellets are a major type of plastics pollution found on shorelines worldwide.This study
investigates the distribution and characteristics of plasticpellets accumulatedon beaches of the Laurentian Great
Lakes of North America and provides a “snapshot”of pellet distribution in a lakesystem that accounts for 21% of
the world's freshwater reserves. We sampled pellets simultaneously from 10m
2
quadrats on 66 beaches and
characterized the 12,595 pellets collected (average of 19.1 pellets/m
2
). Forty-two beaches contained pellets
and 86% of the pellets were found on three beaches: Rossport (Lake Superior), Baxter (Lake Huron), and Bronte
(Lake Ontario). The numberof pellets on each beach was compared withfactors hypothesized to control their ac-
cumulation. In general, positive correlations were found between pellet abundance and watershed population,
number of plastic-related industries, and proximity to a river mouth, although for Lake Superior, abundance
was related to a train spill that tookplace over 10 years ago. Beach grain size appears to be related to pellet abun-
dance, withvery fine sand, fine sand andmedium sand containingthe greatest numberof pellets. All pellets were
visually characterized based on size, color, shape, weathering, and distinguishing traits. The predominant color
was white, oblate shapes were most common, and the main distinguishing trait was a dimple. Most pellets
showed little evidence of weathering, with the weathered samples mainly from Lakes Erie and Ontario. Lake On-
tario pellets were the most varied, with 6/7 shapes, 35/40 colors, and 21/25 distinguishing traits, indicating a
wider range of pellet sources compared to the other lakes. Polymer compositions were mainly polyethylene
Science of the Total Environment 747 (2020) 141227
⁎Corresponding author.
E-mail address: pcorcor@uwo.ca (P.L. Corcoran).
https://doi.org/10.1016/j.scitotenv.2020.141227
0048-9697/Crown Copyright © 2020 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Contents lists available at ScienceDirect
Science of the Total Environment
journal homepage: www.elsevier.com/locate/scitotenv
(PE) and polypropylene (PP). Our results will lead to increased recognition of regional pellet pollution in the
Great Lakes watershed, thereby motivating change during their production, transport and use.
Crown Copyright © 2020 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
1. Introduction
The global proliferation of plastic debris has been widely docu-
mented in both aquatic and terrestrial environments. The first thermo-
sets, synthetic materials that melt when heated and solidify when
cooled, were developed in the late 19th and first half of the 20th centu-
ries. Of all thermoplastics produced, polyethylene (PE) and polypropyl-
ene (PP) are the most common (Andrady and Neal, 2009). Both PE and
PP are derived through polymerization of hydrocarbon monomers in
the presence of heat and a catalyst. Once formed, the plastic is cut into
small particles by a pelletizer and the pellets (also known as “nurdles”)
are then transportedto industries to be melted and extruded or molded
into common plastic products. Unfortunately, numerous pellets are lost
during production, transport and storage (Karlsson et al., 2018). Spilled
or discarded pellets make their way into drainage systems, circulate
through the surface waters of rivers, lakes, seas, and oceans, and are
eventually deposited along shorelines. Documenting, quantifying, and
characterizing pellets provides information that can potentially be
used to change policy or industry behavior in favour of the thoughtful
handling of these materials.
Plastic pellets have been reported from beaches across the globe for
over 45 years (e.g. Carpenter and Smith, 1972;Gregory, 1977;
Zbyszewski et al., 2014;Fernandino et al., 2015;Karlsson et al., 2018).
Identification of pellets thousands of kilometres from the nearest pellet
production or use facility (e.g. Corcoran et al., 2014) indicates that pellet
buoyancy results in long-range transport. If a pellet is encrusted with
minute biota (e.g. Carpenter et al., 1972), the potential for introduction
of invasive species is increased. In addition, pellets have been shown to
adsorb and release persistent organic pollutants (POPs) such as
polychlorinated biphenyls (PCBs), dichloro-diphenyltrichloroethane
(DDT), hexachlorocyclohexanes (HCHs), and polycyclic aromatic hy-
drocarbons (PAHs) (e.g. Mato et al., 2001;Rios et al., 2007;Ogata
et al., 2009;Heskett et al., 2012;Koelmans et al., 2016). International
Pellet Watch, initiated in 2005 by Hideshige Takada, shows that pellets
have been found in over 50 countries, proving that pellet pollution is
truly a globalissue (http://www.pelletwatch.org/index.html). Similarly,
The Great Nurdle Hunt, organized by UK charity FIDRA, aggregates
citizen-collected data on pellets. As of the time of writing, over 3000
“Nurdle Hunts”have taken place, with at least one on every continent
(https://www.nurdlehunt.org.uk/). Once in the environment, pellets
can be ingested by aquatic wildlife such as ocean fish, squid, and sea-
birds (e.g. Braid et al., 2012;Van Franeker and Lavender Law, 2015;
Miranda and de Carvalho-Souza, 2016). Although few studies focus on
the physiological effects of pellet ingestion, numerous investigations
show that microplastic (<5 mm plastic particles) ingestion can affect
feeding behavior, reproduction, and growth of aquatic organisms
(Chae and An, 2017).
Pollution of the Laurentian Great Lakes with plastic debris was first
documented in 2011 in a study examining plastic particle distribution
and degradation along the shoreline of Lake Huron (Zbyszewski and
Corcoran, 2011). Subsequent investigations focused on the presence of
micro- and macroplastic debris in surface waters of the Great Lakes
and its tributaries (Eriksen et al., 2013;Baldwin et al., 2016;Lenaker
et al., 2019), shorelines (Hoellein et al., 2014;Zbyszewski et al., 2014;
Driedger et al., 2015), and benthic lake and river sediment (Corcoran
et al., 2015;Ballent et al., 2016;Dean et al., 2018;Corcoran et al.,
2020). Plastic pellets were components of the debris load in surface
water and beach studies, with very minor amounts reported from ben-
thic zones, as a result of their low density and high surface area.
Zbyszewski and Corcoran (2011) and Corcoran et al. (2015) showed
that pellets are major components of shorelines in the Great Lakes,
with as many as 33 pellets/m
2
(Baxter Beach, Lake Huron) and 21 pel-
lets/m
2
(Humber Bay), respectively.
The overall objective of this study was to record a “snapshot”of the
distribution of plastic pellets along shorelines of the five Laurentian
Great Lakes of North America; a lake system that may be considered a
smaller-scale, freshwater proxy for the world's oceans. The Great
Lakes have approximately 16,500 km of shoreline and hold 21% of the
world's surface freshwater reserves. An essential freshwater resource,
they are simultaneously the location of intense agricultural and indus-
trial activity, including significant plastics manufacturing. Understand-
ing how pellets enter and are transported and distributed through the
lake system is essential in the reduction of pellet pollution.
In this study, we aim to determine the relationships between the
quantity, types and distribution of pellets in each lake to ascertain:
1) the influence of plastic industries and watershed population, 2) the
influence of spatial location with respect to proximity of river mouths
and major highways, and location within or outside bays, and 3) the in-
fluence of depositional processes as related to beachgrain size. By statis-
tically examining these relationships, we convey the extent of pellet
pollution and its possible sources. Through this work, we strive to
raise awareness of the plastic pellet issue and to inspire policy develop-
ment in terms of proper handling of pellets throughout the supply
chain.
2. Materials and methods
2.1. Regional setting
The Laurentian Great Lakes are located in central North America,
straddling the boundary between Canada and the United States
(Fig. 1a). Covering an area of 244,000 km
2
, the Great Lakes system ac-
counts for approximately 21% of the world's surface fresh water and is
home to >30 million people (EPA, 2019). It is composed of the five
main lakes Superior, Michigan, Huron, Erie, and Ontario. Each lake has
its own sub-basins (https://www.ngdc.noaa.gov/mgg/image/images/
greatlakesbasin.pdf) and unique water current patterns (Beletsky
et al., 1999;EPA, 2019).
2.2. Field sampling
A total of sixty-six beaches were surveyed for plastic pellets between
October 7th and 21st in 2018. The number of sampling locationsaround
each lake varied with lake size. Eighteen beaches were sampled on Lake
Superior, 15 on Lake Michigan, 14 on Lake Huron, 10 on Lake Erie, and 9
on Lake Ontario (Fig. 1b). Each lake was surveyed by two people, except
Lake Superior where two people surveyed 15 beaches and three other
people surveyed the remaining 3 beaches. In total, thirteen samplers
were involved in the fieldwork, all were given detailed instructions
prior to departure, and all were familiar with the appearance of plastic
pellets. At each beach, samplers were instructed to i) take photographs
of the beach, strandline, and any plastic debris identified, ii) measure
the grain size of the natural sediment on the beach; if this proved diffi-
cult, samplers collected sediment and brought it backto the lab for mea-
surement, iii) collect all plastic debris along the strandline. The latter
step was conducted bystretchinga 10 m measuring tape perpendicular
to the strandline and collecting pellets and other plastic debris within
1 m of the tape (Fig. 2a). Plastics from only the top 5 cm of the beach
2P.L. Corcoran et al. / Science of the Total Environment 747 (2020) 141227
surface were collected using bare hands if the sediment was wet. Metal
sieves were used if the sediment was dry and fine enough to pass
through the sieve openings, thereby leaving pellets and plastic debris
>2.5 × 3 mm remaining in the sieve. Pellets and other plastic debris
were stored in paper bags, and each bag was labelled according to
lake and beach name. The paper bags were brought back to the Univer-
sity of Western Ontario for pellet characterization.
2.3. Pellet characterization
A total of 508 pellets were randomly selected for chemical analysis
using Fourier transform infrared spectroscopy (FTIR) at Surface Science
Western, University of Western Ontario. Pellet surfaces were analyzed
using platinum attenuated total reflectance (Pt-ATR) equipped with a
diamond crystal in the main box (Bruker Tensor II spectrometer). This
experimental setup allowed for analysis of an area of approximately
2mm×2mmtoadepthof0.6–5μm. The spectra were collected
from 4000 to 400 cm
−1
with a resolution of 4 cm
−1
. Some of thepellets
were also analyzed in cross-section. A total of 101, 99, 105, 100, and 103
pellets were analyzed from Lakes Superior, Erie, Huron, Michigan, and
Ontario, respectively.
Over the course of six months, the pellets were separated from other
plastic items and each pellet was characterized according to size, shape,
diagnostic trait, visible weathering, and color (Fig. 2b–f). Information
concerning each pellet was entered into a database for statistical analy-
sis. The size of each pellet was measured with a ruler by finding the
plane of maximum projection, then measuring the long (l) and interme-
diate (i) axes and multiplying them together. The short axis (thickness)
of each pellet was not measured due to time constraints. The shape of
each pellet was categorized as either circular (l = i and pellet has no an-
gular edges), square (l = i and pellet has four angular edges), rectangu-
lar (l > i and has four angular edges), oblate (l > i and pellet has no
angular edges), cylindrical (l > i; pellet is a circular prism), cylindrical
irregular (l > i and pellet is not similar on all edges of the circular
prism), or irregular (l > i and pellet has no well-defined shape). Diag-
nostic traits (e.g. rims, lines, dimples) were visually determined and di-
vided into23 categories. Color was visually determined and divided into
38 categories. Because six people were involved in pellet characteriza-
tion, the categories for shape and diagnostic traits were cross-checked
among examiners. Color was not cross-checked and therefore repre-
sents the greatest source of characterization error given that different
individuals see color in different ways.
2.4. Spatial analysis
The number of plastic suppliers, distributors and users were deter-
mined from ThomasNet Supplier Discovery and Google Maps using
the search terms “plastics”and “polymers”within a 100–300 km radius
of each lake depending on the size of each lake basin. Coordinates were
assigned to each facility for mapping purposes.
Mapping and spatial analysis for Fig. 7 was performed using ArcMap
10.5. Spatial analysis methods for determining population density by
watershed were modified from Ballent (2016). Dissemination block
area boundary files were provided from Statistics Canada (2020) and
were joined to 2016 census data collected from the Canadian Census
Analyzer (2014). Dissemination block data were used for Ontario. The
profile variables selected from the population and dwelling counts for
2016 included previously normalized data, population density per
square kilometer. United States Census Bureau block-level data were
used from the 2010 US Census for each of the eight states in the Great
Lakes Basin. Watersheds were mapped at the USGS HUC-8 and Ontario
Tertiary levels, which are equivalent (Neff et al., 2005). Census data
were clipped to watersheds and were converted to raster using 100 m
cell size. Zonal statistics were used to determine the population per wa-
tershed, and watersheds were mapped using approximate quantiles.
“Direct Drainage”areas, which indicate islands and land adjacent to
the Great Lakes that were not incorporated into the watershed
shapefiles were not considered in population analysis.
2.5. Statistical analysis
In order to determine whether pellet distribution was related to
river mouth proximity, any beach within 5 km of a river was given a
“true”value, with beaches >5 km from a river mouth assigned a
“false”value for statistical analysis. Similarly, any beach sampled that
was located in a bay was given a “true”value, as opposed to beaches
outside bays, which were given “false”values. Proximity to major high-
ways was determined using the Trans-Canada and 400-series (Canada)
Fig. 1. Spatialand pellet abundance details.A) Location of the Great Lakes system in NorthAmerica. B) Sixty-six sampling locationsalong Great Lakesbeaches, with relative abundances of
pellets indicated.
3P.L. Corcoran et al. / Science of the Total Environment 747 (2020) 141227
and Interstates (U.S.), and distances ranged from 1 to270 km. Figs. 2 to 6
were created in R version 3.6.0., using the package ggplot2 version 3.2.0
(Wickham, 2016). Fig. 1 was created using the package ggmap, version
3.0.0 (Kahle and Wickham, 2013).
3. Results
The total number of pellets collected from around the Great
Lakes was 12,595, for an overall average of 19.1 pellets/m
2
. A total
of twenty-four beach sampling quadrats contained no pellets
(Fig. 1b). The total number of pellets found on Lake Superior
beaches was 1341, for an average of 7.5 pellets/m
2
; 67% of the
beaches contained pellets. On Lake Michigan beaches, a total of
728 pellets were collected, for an average of 4.9 pellets/m
2
,with
73% of the beaches containing pellets. A total of 7471 pellets were
foundonLakeHuronbeacheswithanaverageof57.5pellets/m
2
,
but only 43% of the beaches contained pellets. Lake Erie beaches
contained 302 pellets, for an average of 3.0 pellets/m
2
;80%ofthe
beaches contained pellets. Finally, Lake Ontario beaches contained
2753 pellets with an average of 30.6 pellets/m
2
,butonly56%of
the beaches contained pellets. The overall pellet counts from
Lakes Huron, Ontario and Superior beaches are skewed as a result
of one beach on each lake accounting for 73% (Rossport –Lake Supe-
rior), 96% (Bronte –Lake Ontario), and 97% (Baxter –Lake Huron) of
the pellets (Fig. 3). Removal of these outliers indicates that the
beaches sampled on Lake Michigan contained the greatest number
of pellets, followed by beaches on Lakes Erie, Superior, Huron, and
then Ontario (Fig. S1).
Fig. 2. Sampling site and pellet characteristics. A) Sampling pellets from a 1 × 10 m quadrat along the strandline of a Lake Superior beach. B) Examples of three distinct types of pellets.
Clockwise from top left: black, oblate; bl ack, cylindrical; black, oblate, rimmed. Note that the three types are of different sizes. C) Examples of the three most common shapes,
clockwise from upper left: two circular pellets (rifle beads), two oblate, and three cylindrical. D) and E) Distinguishing traits of pellets sampled. Clockwise from top left in D: rimmed
and nodule; dimple and dirty; rough and broken; rimmed and hole; rimmed. Clockwise from top left in E: groove; dimple and nodule; lines and dirty; lines and dirty. F) Weathered
pellets with distinguishing traits such as dirty, rough and broken. The two lower left pellets are coalesced (joined together).
4P.L. Corcoran et al. / Science of the Total Environment 747 (2020) 141227
Lake
Lake Superior
Lake Michigan
Lake Huron
Lake Erie
Lake Ontario
Number of Pellets
Number of Pellets
Crerar
Escanaba
Grandview
Hilton
Houghton
Humber Bay
Huron Beach
Lakeview Park
Manistique
Marathon
Ontonagon
Pigeon River
Pinconning
Port Austin
Port Clinton
Porter on the Lake
Presque Isle
Providence Bay
Shuniah
Temperance River
Traverse City
Ontario Park
Batchawana Bay
Brevort
Bradford
Cedar Beach
Pentwater
Presqu’ile
Sherman Park
Michipicoten
Park Point
Marquette
Ogden Dunes
Evangola Park
Port Stanley
Harbor Beach
Grand Marais Dunes
Grand Marais
Green Bay
Munising
Holiday Harbour
Private Pfeiffer
South Beach
Algoma
Colchester Beach
Crescent Beach
Brimley
Baraga
Kenosha
Warren Dunes
Sunset Beach
Sodus Point
Katherine Cove
Holland State
Fairport Harbor
Bay View
Wasaga Beach
Colonel Samuel
James N Allen Park
Point Clark
Sheboygan
Bay City
12th Street
Beach Name
Crerar
Escanaba
Grandview
Hilton
Houghton
Humber Bay
Huron Beach
Lakeview Park
Manistique
Marathon
Ontonagon
Pigeon River
Pinconning
Port Austin
Port Clinton
Porter on the Lake
Presque Isle
Providence Bay
Shuniah
Temperance River
Traverse City
Ontario Park
Batchawana Bay
Brevort
Bradford
Cedar Beach
Pentwater
Presqu’ile
Sherman Park
Michipicoten
Park Point
Marquette
Ogden Dunes
Evangola Park
Port Stanley
Harbor Beach
Grand Marais Dunes
Grand Marais
Green Bay
Munising
Holiday Harbour
Private Pfeiffer
South Beach
Algoma
Colchester Beach
Crescent Beach
Brimley
Baraga
Kenosha
Warren Dunes
Sunset Beach
Sodus Point
Katherine Cove
Holland State
Fairport Harbor
Bay View
Wasaga Beach
Colonel Samuel
James N Allen Park
Point Clark
Sheboygan
Bay City
12th Street
Rossport
Baxter Beach
Bronte Beach
Beach Name
Fig. 3. Pellet abundance per sampling quadrat on each beach. Lower graph includes all beaches and inset includes only those beaches with 200 pellets or less. Twenty-four beaches contained no pellets. Bars are color-coded according to lake. (For
interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
5P.L. Corcoran et al. / Science of the Total Environment 747 (2020) 141227
3.1. Pellet compositions
The FTIR results show that 85.8% of the pelletsare composed of poly-
ethylene (PE), 8.5% are polypropylene (PP), 1.6% are thermoplastic
olefin (TPO), and 1.0% are ethylene vinyl acetate (EVA) (Table S1). Poly-
styrene (PS), butadiene-styrene, acrylonitrile butadiene styrene (ABS),
nylon, PP/PE blend, and PE/Polyamide blend make up <2.2% of the pel-
lets. Two pellets were unknown polymers and 2 were minerals. Lakes
A
B
square
rectangular
oblate
irregular
cylindrical irregular
cylindrical
circular
ShapeDistinguishing Traits
Number of Pellets
Lake
Lake Superior
Lake Michigan
Lake Huron
Lake Erie
Lake Ontario
Number of Pellets
white bubble inside
warped
tail
streaks
squished
rough
rimmed
ribbed
other
nodule
lines
line
hollow
hole
groove
dirty
dimple, rimmed
dimple, nodule
dimple, seam
dimple
cut
crack
coalesced
bubbles
broken
Lake
Lake Superior
Lake Michigan
Lake Huron
Lake Erie
Lake Ontario
Fig. 4. Bar graphs illustrating the shapes and distinguishing traits of pellets found on the beaches of each lake. A) The most common shape found on all lakes was oblate, followed by
cylindrical, then circular. B) The most common distinguishing trait found on all lakes was a dimple, although 74% of the pellets contained no distinguishing trait at all.
6P.L. Corcoran et al. / Science of the Total Environment 747 (2020) 141227
yellow white
yellow pink
yellow orange
yellow grey
yellow green
yellow beige
yellow
white red
white brown
white
red
purple blue
purple
pink brown
pink
orange white
orange red
orange grey
orange
grey white
grey purple
grey pink
grey orange
grey
green white
green grey
green brown
green
cyan purple
cyan
brown red
brown orange
brown grey
brown
blue grey
blue beige
blue
black
beige grey
beige
Color
Number of Pellets
Lake
Lake Superior
Lake Michigan
Lake Huron
Lake Erie
Lake Ontario
Proportion of Pellets
yellow white
yellow pink
yellow orange
yellow grey
yellow green
yellow beige
yellow
white red
white brown
white
red
purple blue
purple
pink brown
pink
orange white
orange red
orange grey
orange
grey white
grey purple
grey pink
grey orange
grey
green white
green grey
green brown
green
cyan purple
cyan
brown red
brown orange
brown grey
brown
blue grey
blue beige
blue
black
beige grey
beige
Color
A
B
Fig. 5. Graphsillustratingthe distributionof pellet colors foreach lake. A) The main pelletcolor found on beachesof all lakes was white.The second most commoncolor was black,although
black pellets were notfound on Lake Superiorbeaches. B) Lake Ontario beaches contained the greatest varietyof pellet colors with35/40 colors, whereas the lakewith the poorest variety
of pellet colors was Superior with only 5/40 colors. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
7P.L. Corcoran et al. / Science of the Total Environment 747 (2020) 141227
Erie and Ontario contained the greatest variety of polymer composi-
tions, whereas the pellets analyzed from Lake Superior were composed
of only PE and PP.
3.2. Pellet characteristics
Pellet types on each beach were assigned to pellets of the same size,
and with the same color, shape, diagnostic feature, and visible surface
weathering (Table S2; Fig. 2b). Pellet types on different beaches within
the same lake were compared only for Lake Superior. The sheer volume
of different pellets on beaches of the other four lakes prohibited type
comparisons for this part of the study. Pellet sizes varied between 2
and 42 mm
2
, with a median of 12 mm
2
and a mean of 13.7 mm
2
overall.
Mean sizes for each lake were 17.9 (Superior), 14.5 (Michigan), 13.4
(Huron), 14.5 (Erie) and11.9 (Ontario). A one-way analysis of variance
(ANOVA) was conducted to compare the lake effect on individual pellet
size, that is, to test that there is a significant difference in mean pellet
size between lakes. This was significant at the p < 0.01 level. The pre-
dominant pellet shape overall was oblate (Fig. 4a). Cylindrical, oblate,
and circular pellets were the most common shapes and were identified
in samples from beaches on all five Great Lakes (Fig. 2c). A total of 9329
pellets had no diagnostic traits. Of the other 3266 pellets, dimples were
by far the most common (Lakes Ontario, Huron, Michigan and Supe-
rior), followed by lines (Lakes Ontario and Huron) and nodules (Lake
Ontario) (Figs. 2d, e, 4b). The sampled quadrats on the three beaches
with the greatest number of pellets show that Baxter Beach (Huron)
and Bronte Beach (Ontario) contained the largest variety of pellets
with respect to distinguishing traits, whereas Rossport beach (Superior)
mainly contained pellets with no distinguishing traits.
Pellet color, although a subjective category, shows that the majority
of the pellets on beaches in the Great Lakes Basin were white, followed
by black and grey (Fig. 5a). Lake Ontario, and specifically Bronte Beach,
contained the greatest variety of pellet colors, with thirty-five of the
forty color types (Fig. 5b). Pellets from Lake Huron and Lake Michigan
contained thirteen of forty colors, twelve of forty colors were repre-
sented in Lake Erie pellets, and Lake Superior pellets were of only four
color types (Fig. 5b). A total of 96.4% of the pellets on Lake Superior
beaches were white, oblate, 18 mm
2
and with no diagnostic traits
(Type 3405in Table S2). These pellets are the result of a train derailment
that occurred in 2008 near Terrace Bay, Ontario, which spilled pellets
into Lake Superior (https://www.sootoday.com/local-news/pellets-
mystery-solved-182329).We also documented the relative degree of
weathering of each of 436 PE pelletsanalyzed by FTIR. Peaks in the spec-
tra between 1710 and 1775 cm
−1
are increased absorption peaks,
which are indicative of oxidized material. Oxidation was classified as
low if there was little to no evidence of absorption peaks at about
1715 cm
−1
relative to the characteristic PE peak height at around
1471 cm
−1
(Fig. S2). The results in Table S1 indicate that 13.7% of the
pellets analyzed were weathered, with the greatest percentage of
weathered pellets in Lakes Ontario, Michigan and Erie. We also
attempted to visually note the number of pellets that appeared exten-
sively weathered, as determined through color change (fading or
yellowing), increase in surface roughness (e.g. pits and microfractures),
embrittlement, and abundance of external particles adhered to surfaces
(e.g. Zbyszewski and Corcoran, 2011;Brandon et al., 2016;Cai et al.,
2018)(Fig. 2f). Of the 15,595 pellets examined, 2.4% showed extensive
evidence of weathering (Table S3). By individual lake, the percentages
of pellets that were visually identified as weathered were <0.01% (Su-
perior), <0.01% (Michigan), 2% (Huron), 10.6% (Erie), and 0.04% (On-
tario). A comparison of the results indicates that visual
characterization of weathering is not as precise as chemical identifica-
tion. This may be due to challenges in recognizing discoloration in pel-
lets with dark colors (e.g. black, blue).
The numberof pellets in each beach quadratwas related to grainsize
of the beach sediment. All samplers used the grain size classification
chart of Wentworth (1922), except for the very fine silt tocoarse silt cat-
egories, which were grouped into “silt”because only two beaches fell
into the category (Fig. 6). The “mixed”category was added to indicate
beaches with polymodal grain sizes (containing more than two), and
the “organics”category was added for the Pigeon River beach on Lake
Superior, which was completely covered in logs, sticks and leaves. The
results in Fig. 6 indicate that the majority of pellets were identified on
beaches composed of very fine sand, fine sand, medium sand and
mixed grain populations. These grain size categories also showed the
Grain Size
Organics
Pebbles + Cobbles
Granules
Very coarse sand
Coarse sand + Cobbles
Coarse sand
Mixed
Medium sand
Fine sand
Very fine sand
Silt
Number of Pellets
Fig. 6. Boxplots displaying the relationship between pellet abundance and grainsize on beaches. The means forvery fine sand, fine sand, and medium sand were greaterthan other grain
size grades, which suggests that pellets preferentially accumulate on sandy beaches compared to silty, granular, pebbly and cobble beaches.
8P.L. Corcoran et al. / Science of the Total Environment 747 (2020) 141227
greatest ranges in number of pellets, especially fine sand and medium
sand. Therefore, although pellets appear to accumulate preferentially
on sandy beaches, some sandy beaches contained no pellets.
3.3. Population and plastic industry
High pellet abundances across the Great Lakes were correlated with
large population numbers (Fig. 7). We recognize that the general popu-
lation does not have access to pre-production plastic pellets, however,
distinguishing between industry proximity and watershed population
is challenging because most pellet manufacturers, suppliers and distrib-
utors are located in high population areas (Fig. 7). An exception is Lake
Superior, which has the lowest overall watershed populations of all the
Great Lakes, with only four of the twenty-five watersheds containing
>50,000 people. Removal of the high pellet count at Rossport beach,
which is a direct result of the train derailment, leaves a total of 368 pel-
lets (2.0 pellets/m
2
). Furthermore, removal of all pellets from Lake Su-
perior beaches that originated from the spill (Type 3405 in Table S2)
leaves only 48 pellets (0.3 pellets/m
2
).
Lake Michigan pellet counts appear to be greater with an increased
number of plastic industries, particularly in the southeastern and south-
western portions of the lake (Fig. 7). The three northernmost sampling
locations and two southernmost sampling locations contained a mean
of 2 pellets(0.2 pellets/m
2
) and these were the regions with the lowest
number of industries.
Beaches along the northern half of Lake Huron contained no pellets
and twenty-one of those twenty-four watersheds had total populations
<100,000 (Fig. 7). In contrast, the southern half of Lake Huron (includ-
ing the southern coastline of Georgian Bay) is composed of twenty wa-
tersheds, nine of which have a population>100,000; pellet abundances
are substantially greater in the southern (67.1 pellets/m
2
)comparedto
the northern (0 pellets/m
2
) parts of the lake (Fig. 7).In particular, the lo-
cation with the highest pellet counts across all five Great Lakes is in the
Sarnia region at the southern end of the lake. Although most of the plas-
tic industries in this region are found along the St. Clair River, which
flows south into Lake St. Clair, several creeksdraining the industrial sec-
tor flow into southern Lake Huron. We have visually identified numer-
ous pellets within and along at least one of these creeks. Although
pellets arewidely distributed across LakeErie, there appears to be no re-
lationship between pellet abundance and the location and density of
plastic industries (Fig. 7). Notwithstanding, the great variety in pellet
compositions and colors for Lake Erie samples compared to other
lakes supports the hypothesis that multiple industrial sources are
involved.
Of the twenty-two watersheds surrounding Lake Ontario, sixteen
have watershed populations >100,000 and two watersheds contain
Fig. 7. GISmap displaying pellettotals at each beach(yellow dots), and locations ofplastic distributors, manufacturers,and suppliers throughout the entire Great Lakes basin.Watersheds
are divided by thick black lines and total population per watershed is indicated in various shades of grey. Note how the pellet totals are greatest in Lake Huron and Lake Ontario. (For
interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
9P.L. Corcoran et al. / Science of the Total Environment 747 (2020) 141227
>1,000,000 people. The high pellet count of 26.4 pellets/m
2
at Bronte
Beach falls within one of these watersheds (Fig. 7). Bronte Beach is lo-
cated within an approximately 160 km long corridor spanning two wa-
tersheds (Credit-16 mile, Humber-Don) that contain 415 identified
plastic industries (Fig. 7).
3.4. Location relative to bays, river mouths and major highways
The number of pellets at each sampling quadrat was compared with
proximity to rivers, location within bays, and proximity to major high-
ways. Thirty-one of the sixty-six beaches sampled are located within
5 km of a river, and these contained a mean number of 373 pellets
(Table S2). Notably, the three beaches with the highest pellet counts
(Baxter, Bronte, and Rossport) are all located within 5 km of a river.
For sampling locations with <200 pellets, the mean was 36. The
thirty-five beaches located >5 km from a river had a mean of 20 pellets.
Atwo-sample,one-tailedt-test on the data containing all locations pro-
duces a p-value of 0.0747, indicating that there is a significantly greater
amount of pellets on beaches within 5 km of a river mouth.
Of all of the beaches sampled, seventeen were located within a bay
and the mean number of pellets for these beaches was 21 (Table S2).
The 49 beaches located outside bays produced a mean pellet abundance
of 240. For locations with <200 pellets, the means for pellet abundance
within bays (21) and outside bays (29) were not statistically significant.
A simple two-sample t-test on thedata containingall locations produces
a p-value of 0.4214, indicating that there is no significant difference in
mean number of pellets on beaches within and outside bays. The num-
ber of pellets at each sampling location was also plotted against distance
to a major highway. The data points display no correlation between the
two variables (Fig. S3).
4. Discussion
The main objective of this study was to determine the factors con-
trolling the abundance and distribution of plastic pellets along beaches
throughout the Great Lakes watershed. The results of this study high-
light some of the major controls on pellet distribution during late Au-
tumn, 2018.
4.1. Influence of plastic industries and watershed population
Not surprisingly, the data show that in the basin as awhole, high pel-
let abundance can be related to high watershed population and greater
numbers of plastic industries, as the two factors are positively corre-
lated. The two beaches with the greatest number of pellets (Bronte-
Lake Ontario; Baxter-Lake Huron) are spatially associated with a signif-
icant number of plastic industries, indicating their influence on pellet
abundance. Lake Superior contained the beach with the third highest
number of pellets (Rossport), but this abundance is neither due to wa-
tershed population nor industry, as both factors are relatively negligible
for this lake. Instead, the pellets on Lake Superior beaches are a result of
a spill from a train derailment, indicating that pellets are not only lost at
the source or destination, but also in transit.
Interestingly, Lake Erie beaches contained the lowest number ofpel-
lets, but displayed a wide pellet distribution across its beaches.The plas-
tic debris distribution models of Hoffman and Hittinger (2017) and
Cable et al. (2017) suggest that most plastic pollution in Lake Erie
would accumulate in the water along the southern shoreline. However,
we did not see a correlation between our beached pellet data and the
surface water accumulation models. The lack of a statistical relationship
between pellet abundance, watershed population, and number of plas-
tic industries for Lake Erie may be a function of its short hydraulic resi-
dence (retention) time. According to the EPA (2019), Lake Erie has a
residence time of approximately 2.6 years, whereas in contrast,the res-
idence time for Lake Superior is 191 years.
There does appear to be a correlation between our pellet data and
Hoffman and Hittinger's (2017) modelled movement of plastic debris
for Lake Michigan. Both data sets show a greater accumulation of plastic
debris in the southern portion of the lake. Unlike Lake Erie, Lake Michi-
gan has a high hydraulic residence time of 99 years (EPA, 2019), and
therefore, residence time does not appear to control the low number
of pellets on Lake Michigan beaches overall. The low pellet numbers, es-
pecially for the high population watershed in which 12th Street beach
was sampled (with 0 pellets found), and wide distribution of pellets
across beaches of Lake Michigan may be a result of: i) the drainage of
Chicago's land into the Mississippi River system instead of Lake Michi-
gan through a complex natural and artificial hydrological network
known as the Chicago Area Waterway System (CAWS) (Duncker and
Johnson, 2016), ii) a low overall release/spill of pellets, or iii) the pellets
being retained in surface water and not deposited on the beaches. Com-
paring the movement of surface water currents within each lake is be-
yond the scope of this project, but retention of pellets in surface
waters may be a function of the complexity of the current patterns
(e.g. number of gyres, current strength and speed) in different seasons
of the year.
Interestingly, there were statistically significant differences in pellet
sizes between lakes. The largest pellets were identified on Lake Superior
and the smallest were found on Lake Ontario. We considered that this
could be related to a decrease in size with extended weathering asa pel-
let travels through the Great Lakes system, but the results show that
Lake Erie beaches contained the most highly weathered pellets and
their average size of 14.5 mm
2
is equivalent and greater than the size
averages for Lake Michigan and Lake Huron, respectively. Instead, the
average size differences, much like color and shape, probably reflect
the producers' preferences, and have little to do with environmental
effects.
We investigated the characteristics of each pellet in order to identify
different pellet types on each beach, which would help find clues
concerning source. The number of different pellet types on the most
pellet-rich beach of each lake included: 6/973 (Rossport-Superior),
426/7268 (Baxter-Huron), 79/167 (Sheboygan-Michigan), 58/127
(Fairport Harbor-Erie), and 1291/2635 (Bronte-Ontario). These results
suggest that there is an order of magnitude fewer sources of pellets to
Lake Superior than to Lake Huron, and that Lake Huron contains an
order of magnitude fewer pellet sources than Lakes Ontario, Erie and
Michigan. Lake Ontario, and specifically Bronte Beach, contained the
greatest number of pellet types per total pellets. If plastic industries
are purposefully or inadvertently allowing pellets to spill into or near
tributaries in the Credit-16 mile and Humber-Don watersheds (which
flow into Lake Ontario), this could explain the wide range of pellet
types washing up on Bronte Beach. While large, instantaneous spills
have required cleanups (e.g. Terrace Bay, ON and Pocono Creek, PA), in-
dustries responsible for long-term, regular releases of pellets have not
had such requirements. However, a recent federal decision sets a new
precedent. For decades, the Formosa Plastics Facility in Point Comfort,
TX discharged plastic pellets and powders into Lavaca Bay despite
their permit allowing only trace amounts of floating debris
(Waterkeeper v. Formosa, June, 2019a). In the Final Consent Decree,
the US District Court required, among other remedial measures, that
past discharges of plastics had to be cleaned up, future discharges
abated, and that Formosa put $50 million USD towards Environmental
Mitigation Projects (Waterkeeper v. Formosa, Nov., 2019b).
4.2. Influence of spatial location (river mouths, bays, highways)
Plastic pellets are normally transported from manufacturer to
processer by train, truck or ship. The pellets spilled from a rail car into
Lake Superior in 2008 continue to be deposited along the lake's shore-
line, but the long hydraulic residence time of the lake suggests that
the pellets may remain suspended in the water column for a century
or more. Plastic pellets spilled during transport via trucks (e.g. https://
10 P.L. Corcoran et al. / Science of the Total Environment 747 (2020) 141227
www.wfmz.com/news/tractor-trailer-spills-plastic-pellets-in-pocono-
creek/article_970503dd-90a8-578c-8a51-bfe2f0214693.html), could
also make their way into drainage systems, creeks and other water-
courses, and eventually become deposited into large bodies of water.
We attempted to discern if a positive relationship exists between pellet
abundance on Great Lakes beaches and proximity to major highways.
The data indicate that no correlation could be found, which may be a
function of distance between source (truck) and sink (lake).
We also hypothesized that once pellets enter a lake, they would
preferentially accumulate inbays or other protected inlets. No statistical
support for this hypothesis wasevident inour data, which suggests that
the low density of pellets causes them to recirculate throughout each
lake, become beached during high onshore wind and wave events,
and then become transported back into the lake during high rain, lake
water or offshore wind events.
One factor that positively affects the abundance of pellets onbeaches
in our study isproximity toriver mouths. A statistically greater number
of pellets were found on beaches located within 5 km of a river mouth.
This supports models showing that rivers are one of the main pathways
transporting plastic debris from land to larger bodies of water (Lebreton
et al., 2017;Schmidt et al., 2017). Studies have also shown that
microplastics, including pellets, are abundant in tributaries flowing
into the Great Lakes watershed (Corcoran et al., 2015;Baldwin et al.,
2016;McCormick et al., 2016;Lenaker et al., 2019), and therefore, trib-
utaries could represent major pathways for input of pellets to the lakes.
4.3. Influence of depositional processes as related to beach grain size
The preferential deposition and/or retention of pellets on beaches
with sand-size sediment is likely a function of waveenergy in the depo-
sitional environment. Beaches composed of grains larger than medium
sand require relatively high energy, and cobble beaches, such as Mara-
thon (Superior) and Porter on the Lake (Ontario) contained no pellets.
Although high wave energy would transport pellets onto the beach,
that same energy redistributes the pellets back into the water. Beaches
composed of silt or organics (e.g. Port Clinton-Huron; Pigeon River-
Superior) also contained no pellets, but this is instead due to the wave
energy being too low for sufficient landward transport. Although it is
expected that grooming affects the distribution of pellets across a
beach, only 6 of the 66 beaches had been groomed prior to sampling
(Table S3), and therefore, we could not statistically test this hypothesis.
5. Conclusions and future work
The present basin-wide investigation of plastic pelletsrepresents the
largest simultaneous sampling campaign for pellets in the world. The
study emphasizes the major controls on pellet pollution in the Lauren-
tian Great Lakes. Factors that correlate positively with elevated pellet to-
tals are high watershed population numbers, high density of plastic
industries, <5 km distance from a river mouth, past evidence of pellet
spills, and beach grain sizes ranging between very fine and medium
sand. The low number of pellet types and compositions on Lake Supe-
rior beaches and the high number of types and compositions on Lake
Ontario beaches support the influence of plastic manufacturers and
processers on pellet pollution. Ideally, pellet types across all five lakes
should be compared in order to determine whether pellets are being
transported throughout the Great Lakes system. This is the next logical,
albeit lengthy extension of the present study. We are planning to survey
each sampling site and collect pellets in October 2023, to compare pellet
abundances from Year 1 and Year 5. The results will hopefully highlight
whether awareness and mitigation of pellet spills has improved. An ad-
ditional next step in this comprehensive investigation is to determine
whether the pellets on Great Lakes beaches have adsorbed persistent
organic pollutants (POPs) on their surfaces. The combination of plastic
debris and harmful pollutants could prove detrimental to aquatic ani-
mals and grazing birds in the various ecosystems of the Great Lakes.
Finally, the identification of shape, size, diagnostic trait, and color
provided in Table S2 could prove very useful for pellet manufacturers
and processers who wish to determine whether their products are con-
tributing to pellet pollution in the Great Lakes watershed.
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.scitotenv.2020.141227.
CRediT authorship contribution statement
Patricia L. Corcoran: Conceptualization, Data curation, Formal anal-
ysis, Methodology, Project administration, Supervision, Visualization,
Writing - original draft. Johanna de Haan Ward: Formal analysis, Soft-
ware, Visualization, Writing - review & editing. Ian A. Arturo: Data
curation,Formal analysis, Investigation, Methodology, Software, Visual-
ization, Writing - review & editing. Sara L. Belontz: Conceptualization,
Data curation, Formal analysis, Investigation, Methodology, Software,
Visualization, Writing - review & editing. Tegan Moore: Conceptualiza-
tion, Data curation, Formal analysis, Investigation, Writing - review &
editing. Carolyn M. Hill-Svehla: Data curation, Formal analysis, Writing
- review & editing. Kirsty Robertson: Conceptualization, Data curation,
Formal analysis, Writing - review& editing. Kelly Wood: Conceptualiza-
tion, Data curation, Formal analysis, Writing - review & editing. Kelly
Jazvac: Conceptualization, Data curation, Funding acquisition, Writing
- review & editing.
Declaration of competing interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence the work reported in this paper.
Acknowledgements
This investigation would not have been possible without help from
numerous individuals. We are grateful to Zhaoming Jiang, Jasmine Nieva,
and Nina Kozikowski for assisting with pellet characterization, and Lorena
Rios Mendoza, Christina Battle, Eeva Siivonen, Daniela Leon Vargas, Chia-
An Lin, and José Avalos for helping with beach surveys. Thank you to
Heather Davis who was involved with conceptualization of the project.
Thanks go out to Doug Woolford and Simon Bonner for their input
concerning statistical methods, and to Elliott Elliott for assisting with in-
dustry searches. Rebecca Sarazen conducted FTIR analysis of the pellets.
This work was supported by SSHRC Insight Grant #435-2017-1253 (Lead
Applicant: Kelly Jazvac). We acknowledge that the fieldwork for this re-
search took place on the traditional territories of the Anishinabewaki,
Haudenosaunee, Huron-Wendat, Attiwonderonk (Neutral), Onondaga,
Odǫhwęja:deˀ(Cayuga), Onöndowa'ga:’(Seneca), Wenrohronon, Erie,
Miami, Peoria, Meškwahki·aša·hina (Fox), Bodéwadmiké (Potawatomi),
oθaakiiwaki‧hina‧ki(Sauk),Odawa,Petun,Metis,Očeti Šakówiŋ(Sioux),
Kiikaapoi (Kickapoo), and Menominee peoples. We are grateful to the In-
digenous peoples who have been protectors of the Great Lakes since time
immemorial.
References
Andrady, A.L., Neal, M.A., 2009. Applications andsocietal benefits of plastics. Phil. Trans.R.
Soc. B 364, 1977–1984.
Baldwin, A.K., Corsi, S.R., Mason, S.A., 2016. Plastic debris in 29 Great Lakes tributaries: re-
lations to watersh ed attributes and hydrology. Environ. Sci. Technol. 50,
10377–10385.
Ballent, A.M., 2016. Anthropogenic Particles in Natural Sediment Sinks: Microplastics Ac-
cumulation in Tributary, Beach and Lake Bottom Sediments of Lake Ontario, North
America. Thesis and Dissertation Repository. , p. 3941. https://ir.lib.uwo.ca/etd/3941.
Ballent, A., Corcoran, P.L., Madden, O., Helm, P.A., Longstaffe, F.J., 2016. Sources and sinks
of microplastics in Canadian Lake Ontario nearshore, tributary and beach sediments.
Mar. Pollut. Bull. 110, 383–395.
Beletsky, D., Saylor, J.H., Schwab, D.J., 1999. Mean circulation in the Great Lakes. J. Great
Lakes Res. 25, 78–93.
11P.L. Corcoran et al. / Science of the Total Environment 747 (2020) 141227
Braid, H.E., Deeds, J., DeGrasse, S.L., Wilson, J.J., Osborne, J., Hanner, R.H., 2012. Preying on
commercial fisheries andaccumulating paralytic shellfishtoxins: a dietary analysisof
invasive Dosidicus gigas (Cephalopoda Ommastrephidae) stranded in PacificCanada.
Mar. Biol. 159, 25–31.
Brandon, J., Goldstein, M., Ohman, M.D., 2016. Long-term aging and degradation of
microplastic particles: comparing in situ oceanic and experimental weathering pat-
terns. Mar. Pollut. Bull. 110, 299–308.
Cable, R.N.,Beletsky, D., Beletsky,R., Wigginton, K., Locke, B.W., Duhaime,M.B., 2017. Dis-
tribution and modeled transport of plastic pollution in the Great Lakes, the world's
largest freshwater resource. Front. Environ. Sci. 5, 1–18.
Cai, L., Wang, J., Peng, J., Wu, Z., Tan, X., 2018. Observation of the degradation of three
types of plastic pellets exposed to UV irradiation in three different environments.
Sci. Total Environ. 628-629, 740–747.
Canadian Census Analyser, 2014. Welcome to the Canadian Census Analyser. Faculty of
Arts and Science, University of Toronto http://datacenter.chass.utoronto.ca/census/
index.html. (Accessed 23 February 2020).
Carpenter, Smith, 1972. Carpenter, E.J., Smith Jr., K.L., 1972. Plastics on the Sargasso Sea
surface. Science 175, 1240–1241.
Carpenter, E.J., Anderson, S.J., Harvey, G.R., Miklas, H.P., Peck, B.B., 1972. Polystyrene
spherules in coastal waters. Science 178, 749–750.
Chae, Y., An, Y.J., 2017. Effects of micro- and nanoplastics on aquatic ecosystems: current
research trends and perspectives. Mar. Pollut. Bull. 124, 624–632.
Corcoran,P.L., Moore, C.J., Jazvac, K., 2014. An anthropogenic marker horizon in the future
rock record. GSA Today 24, 4–8.
Corcoran,P.L., Norris, T., Ceccanese, T., Walzak, M.J., Helm, P.A., Marvin, C.H., 2015. Hidden
plastics of Lake Ontario, Canada and their potential preservation in the sediment re-
cord. Environ. Pollut. 204, 17–25.
Corcoran, P.L., Belontz, S.L., Ryan, K., Walzak, M.J., 2020. Factors controlling the distr ibu-
tion of microplastic particles in benthic sediment of the Thames River, Canada. Envi-
ron. Sci. Technol. 54, 818–825.
Dean, B.Y., Corcoran, P.L., Helm, P.A., 2018. Factors influencing microplasticabundances in
nearshore, tributary and beach sediments along the Ontario shoreline of Lake Erie.
J. Great Lakes Res. 44, 1002–1009.
Driedger, A.G.J., Dürr, H.H., Mitchell, K., Van Cappellen, P., 2015. Plastic debris in the Lau-
rentian Great Lakes: a review. J. Great Lakes Res. 41, 9–19.
Duncker, J.J., Johnson, K.K., 2016. Hydrology of and Current Monitoring Issues for the Chi-
cago Area Waterway System, Northeastern Illinois (ver. 1.1 March 2016): USGS Sci-
entific Investigations Report 2015–5115. 48 p. https://doi.org/10.3133/sir20155115.
EPA (United States Environmental Protection Agency), 2019. Facts and figures about the
Great Lakes. https://www.epa.gov/greatlakes/facts-and-figures-about-great-lakes.
Eriksen, M., Mason, S., Wilson, S., Box, C., Zellers, A., Edwards, W., Farley, H., Amato, S.,
2013. Microplastic pollution in the surface waters of the Laurentian Great Lakes.
Mar. Pollut. Bull. 77, 177–182.
Fernandino, G., Elliff, C.I., Silva, I.R., Bittencourtc, A.C.S.P., 2015. How many pellets are too
many? The pellet pollution index as a tool to assess beach pollution by plastic resin
pellets in Salvador, Bahia, Brazil. J. Integrat. Coast. Manag 15, 325–332.
Gregory, 1977. Plastic pellets on New Zealand beaches. Mar. Pollut. Bull. 8, 82–84.
Heskett, M., Takada, H., Yamashita, R., Yuyama, M., Ito, M., et al., 2012. Measurement of
persistent organic pollutants (POPs) in plastic resin pellets from remote islands: to-
ward establishment of background concentrations for International Pellet Watch.
Mar. Pollut. Bull. 64, 445–448.
Hoellein, T., Rojas, M., Pink, A., Gasior, J., Kelly, J., 2014. Anthropogenic litter in urban
freshwater ecosystems: distribution and microbial interactions. PLoS One 9, e98485.
Hoffman, M.J., Hittinger, E., 2017. Inventory and transport of plastic debris in the Lauren-
tian Great Lakes. Mar. Pollut. Bull. 155, 273–281.
Kahle, D., Wickham, H., 2013. ggmap: spatial visualization with ggplot2. R Journal 5 (1),
144–161.
Karlsson, T.M., Vethaak, A.D., Almroth, B.C., Ariese,F., van Velzen, M., Hassellov, M., Leslie,
A., 2018. Screening for microplastics in sediment, water, marine invertebrates and
fish: method development and microplastic accumulati on. Mar. Pollut. Bull. 122,
403–408.
Koelmans, A.A., Bakir, A., Burton, G.A., Janssen, C.R., 2016. Microplastic as a vector for
chemicals in the aquatic environment: critical review and model-supported reinter-
pretation of empirical studies. Environ. Sci. Technol. 50, 3315–3326.
Lebreton, L.C.M., van der Zwet, J., Damsteeg, J., Slat, B., Andrady, A., Reisser, J., 2017. River
plastic emissions to the world’s oceans. Nat. Commun. 8, 15611.
Lenaker, P.L., Baldwin, A.K., Corsi, S.R., Mason, S.A., Reneau, P.C., Scott, J.W., 2019. Vertical
distribution of microplastics in thewater column and surficialsediment from the Mil-
waukee River Basin to Lake Michigan. Environ. Sci. Technol. 53, 12227–12237.
Mato, Y., Isobe, T., Takada, H., Kanehiro, H., Ohtake, C., Kaminuma, T., 2001. Plastic resin
pellets as a transport medium for toxic chemicals in the marine environment. Envi-
ron. Sci. Technol. 35, 318–324.
McCormick, A.R. , Hoellein, T.J., London, M.G., Hit tie, J., Scott, J.W., Kelly, J.J., 2016.
Microplastic in surface waters of urban rivers: concentration, sources, and associated
bacterial assemblages. Ecosphere 7 (e01556).
Miranda, D. de A., de Carvalho-Souza, G.F., 2016. Are we eating plastic-ingesting fish?
Mar. Pollut. Bull. 103, 109–114.
Neff, B.P., Day, S.M., Piggott, A.R., Fuller, L.M., 2005. Base Flow in the Great Lakes Basin:
USGS Investigations Report 2005-5217. 23 p. https://pubs.usgs.gov/sir/2005/5217/
pdf/SIR2005-5217.pdf.
Ogata, Y., Takada, H., Mizukawa, K., Hirai, H., Iwasa, S., Endo, S., Mato, Y., Saha, M., Okuda,
K., Nakashima, A., Murakami, M., Zurcher, N., Booyatumanondo, R., Zakaria, M.P.,
Dung, L.Q., Gordon, M., Miguez, C., Suzuki, S., Moore, C., Karapanagioti, H.K., Weerts,
S., McClurg, T., Burres, E., Smith, W., Velkenburg, M.V., Lang, J.S., Lang, R.C., Laursen,
D., Danner, B., Stewardson, N., Thompson, R.C., 2009. International pellet watch:
global monitoring of persistent organic pollutants (Pops) in coastal waters. 1. Initial
phase data on PCBs, DDTs, and HCHs. Mar. Pollut. Bull. 58, 1437–1446.
Rios, L.M., Moore, C., Jones, P.R., 2007. Persistent organic pollutants carried by synthetic
polymers in the ocean environment. Mar. Pollut. Bull. 54, 1230–1237.
Schmidt,C., Krauth, T., Wagner, S., 2017. Export of plasticdebris by Rivers into thesea. En-
viron. Sci. Technol. 51, 12246–12253.
Statistics Canada, 2020. Population an d dwelling count highlight table, 2016 census.
https://outlook.office.com/mail/inbox/id/AAQkADk2Y2RkYWZjLTJlYWYtNDkyNy1
iOWM0LTVhNTcxMDJlYTgwNwAQAMfo5c88FRtFmLCmEIX4OeU%3D. (Accessed 21
February 2020).
Van Franeker, J.A., Lavender Law,K., 2015. Seabirds, gyres and global trends in plastic pol-
lution. Environ. Pollut. 203, 89–96.
Waterkeeper v. Formosa PlasticsCorp, 2019a. Texas,No. 6:17-CV-0047.WL 2716544 (S.D.
Tex. June 27, 2019) (Memorandum and Order).
Waterkeeper v. Formosa Plastics Corp, 2019b. Texas, CIVIL ACTION NO. 6:17-CV-47 (S.D.
Tex. November 27) (Final Consent Decree).
Wentworth, C.K., 1922. A scale of grade and class terms for clastic sediments. J. Geol. 30,
377–392.
Wickham, H., 2016. ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag, New
York ISBN 978-3-319-24277-4. https://ggplot2.tidyverse.org.
Zbyszewski, M., Corcoran, P.L., 2011. Distribution and degradation of fresh water plastic
particles along th e beaches of Lake Huron, Canada. Water Air Soil Pollut. 220,
365–372.
Zbyszewski, M., Corcoran, P.L., Hockin, A., 2014. Comparison of the distribution and deg-
radation of plastic debris along shorelines of the Great Lakes, North America. J. Great
Lakes Res. 40, 288–299.
12 P.L. Corcoran et al. / Science of the Total Environment 747 (2020) 141227
