Team MARINE: Microplastic Analysis and Removal in Industrial and Natural Ecosystems

Josh DiGiorgio, Alana Ginsburg, Julia Grafstein, Cameron Hobbs, Lindsey Moore,

Robert Pang, Jonah Pereyra, and Fiona Quin Zabel

Gemstone Research Program, Honors College, The University of Maryland, College Park

Dr. Lance Yonkos

Thesis submitted in partial fulfillment of the requirements of the Gemstone Honors Program,

University of Maryland, 2024

Committee: Dr. Lance Yonkos, Dr. Fred Pinkney, Dr. Rob Hale, Kelly Somers, Dr. Zhi Yang

Soon, Matt Robinson



2

Abstract

Numerous methods have been developed for microplastics isolation and quantification in

various environmental media, many of which require elaborate or expensive analytical

equipment and decontaminated lab space. This study seeks to create a reproducible and

economical method for isolating microplastics in surface water that rely on Nile Red staining.

We use a Nile Red pre-staining step prior to sample digestion, density separation, and filtration to

mitigate downstream in-lab contamination before quantification of microplastics via fluorescent

microscopy. To test the method, we collected replicate surface water samples from several

reaches of an urban stream in the Chesapeake Bay Watershed, USA seasonally for one year. The

proposed sampling and quantification method found some success in these surface water samples

with specific microplastics (≥20 µm) able to be enumerated.



3

Table of Contents

Abstract............................................................................................................................................2
Table of Contents............................................................................................................................. 3
List of Figures................................................................................................................................................ 5
Introduction......................................................................................................................................8
Sampling........................................................................................................................................ 15

Site Selection........................................................................................................................... 15
Collection.................................................................................................................................29

Processing...................................................................................................................................... 35
Acidification and Digestion..................................................................................................... 36
Staining.................................................................................................................................... 38
Density Separation................................................................................................................... 39

Quantification................................................................................................................................ 43
Data and Results............................................................................................................................ 48

Particle Classification Results..................................................................................................48
Seasonality Results.................................................................................................................. 49
Location Results.......................................................................................................................50
ANOVA Statistics.................................................................................................................... 53
ANOVA Assumptions..............................................................................................................53
ANOVA Statistics.................................................................................................................... 53
Interrater Agreeableness.......................................................................................................... 55

Discussion......................................................................................................................................56
Methods Utility and Sources of Contamination.......................................................................56
Comparable Local Study..........................................................................................................62
Effect of Environmental Conditions........................................................................................ 62
Policy Implications.................................................................................................................. 67
Future Research Directions......................................................................................................73

Equity Impact Analysis..................................................................................................................74
Environmental Justice..............................................................................................................74
Accessible Protocols................................................................................................................ 77

Budget Addendum......................................................................................................................... 78
Acknowledgments..........................................................................................................................78
References......................................................................................................................................79
Appendix A....................................................................................................................................91
Methodology Protocols................................................................................................................................ 91
Appendix B.................................................................................................................................... 97



4

Results Supplementary Figures....................................................................................................................97
Appendix C.................................................................................................................................................. 99
Budget Addendum....................................................................................................................................... 99



5

List of Figures & Tables

Figure 1: Diagram of Microplastics Transport Pathways in the Ocean (Ashrafy et al., 2023).......8
Figure 2: Anacostia watershed study site locations...................................................................... 15
Table 1: Site Names, Numbers, and Coordinates..........................................................................16
Figure 3: Bladensburg Waterfront Park Sampling Site.................................................................17
Figure 4: Bladensburg Waterfront Park Sampling Site.................................................................18
Figure 5: Bladensburg Waterfront Park Sampling Site.................................................................18
Figure 6: Bladensburg Waterfront Park Sampling Site.................................................................19
Figure 7: Northeast Branch Anacostia River Stream Gauge Sampling Site.................................20
Figure 8: Northeast Branch Anacostia River Stream Gauge Sampling Site.................................20
Figure 9: Northeast Branch Anacostia River Stream Gauge Sampling Site.................................20
Figure 10: College Park Airport Sampling Site............................................................................21
Figure 11: College Park Airport Sampling Site............................................................................ 21
Figure 12: College Park Airport Sampling Site............................................................................22
Figure 13: Iribe Sampling Site......................................................................................................23
Figure 14: Iribe Sampling Site......................................................................................................23
Figure 15: Iribe Sampling Site......................................................................................................23
Figure 16: Lot 11 Sampling Site...................................................................................................24
Figure 17: Lot 11 Sampling Site...................................................................................................24
Figure 18: Lot 11 Sampling Site...................................................................................................24
Figure 19: Rita’s Sampling Site.................................................................................................... 25
Figure 20: Rita’s Sampling Site.................................................................................................... 26
Figure 21: Rita’s Sampling Site.................................................................................................... 26
Figure 22: Golf Course Sampling Site..........................................................................................27
Figure 23: Golf Course Sampling Site..........................................................................................27
Figure 24: Golf Course Sampling Site..........................................................................................28
Figure 25: Technical Diagram of Sampling Apparatus Showing Dimensions............................. 30
Table 2: Sampling Apparatus Parts List........................................................................................31
Figure 26: CAD Representation of Sampling Apparatus............................................................. 32
Figure 27: Apparatus Being Used to Collect a 1 L Sample..........................................................33
Table 3: Collection Dates and their Numerical Code....................................................................34
Figure 28: Filtration Set-Up..........................................................................................................35
Figure 29: Sediment-Clogged Slides Which did not Receive Density Separation.......................36
Figure 30: Centrifuge used for these analyses.............................................................................. 39
Figure 31: Sample Post-Density Separation................................................................................. 40
Table 4: Guidelines for Our Classification of Different Items that Fluoresce.............................. 44



6

Table 5: Comparison of Fluorescent and Light Microscope View of Different Plastic Shapes and
Types.............................................................................................................................................. 45
Table 6: Descriptive Statistics of Average Sample Particle Distribution per Liter.......................48
Figure 32: Average distribution of particle types across samples.................................................48
Figure 33: Average Total Particles by Season.............................................................................. 49
Figure 34: Average Total Particles by Location............................................................................50
Figure 35: Average Particles by Location and Season..................................................................51
Figure 36: Total Microplastics Count Between Sites and Between Seasons. Error bars represent
the range between the totals of two samples at each location (only one control sample was taken
for each date, so there is insufficient data for error bars on control samples)............................... 51
Table 7: One-Way Anova Results for Significant Differences in Quantities of Particle Sizes
Across Seasons...............................................................................................................................53
Table 8: One-Way Anova Results for Significant Differences in Quantities of Particle Sizes
Across Site Locations.................................................................................................................... 54
Table 9: Two-Way Z-Test for Counter Differentials Significantly Different from Zero.............. 55
Table 10: Microplastics Counts from Lab Control Samples.........................................................56
Figure 37: Microscope Photograph of Stained Potassium Hydroxide Solution........................... 58
Figure 38: Collection of Microscope-view Photographs of Different Types of Plastics. Rows
represent differing processing procedures (Tong et al., 2021).......................................................60
Figure 39: Turbidity on the Days that We Sampled as Reported by the Northeast Regional
Stream Gauge (USGS, 2024).........................................................................................................63
Figure 40: Discharge Volume on the Days that We Sampled as Reported by the Northeast
Regional Stream Gauge (USGS, 2024)......................................................................................... 63
Figure 41: Water temperature on the days that we sampled as reported by the Northeast Regional
Stream Gauge (USGS, 2024).........................................................................................................64
Figure 42: Water Height at both Bladensburg and Northeast Regional Stream Gauge Sites on
Sample Days and Precipitation as Reported at NOAA HYATTSVILLE 0.7 N, MD US station
(NCEI, 2024)..................................................................................................................................64
Table 11: Tidal Positions From Our Sampling Days as Reported by NOAA Anacostia Watershed
Monitoring Site (NOS, 2024)........................................................................................................ 65
Figure 43: Distance from Major Roads between Bladensburg, Golf Course, and Iribe Sampling
Sites................................................................................................................................................72



7

Introduction

Single-use plastics and plastic derivatives have become pervasive throughout the modern

world, with plastic items serving as the de facto standard for cheap and quick-to-manufacture

items with a limited purpose. Microplastics are an unavoidable byproduct of the widespread use

of plastics and as a result, have been detected in nearly every location searched (Cole et al.,

2011). Microplastics are defined as any piece of plastic that is 5mm or less in length, and

nanoplastics form the subcategory of microplastics with sizes ranging between 1 nm and 1 µm

(Materić et al., 2022). These plastics come from a variety of sources, which are generally broken

down into two main categories: primary microplastics and secondary microplastics. Primary

microplastics are those plastics that, due to their intended use, are manufactured as a microplastic

(size less than 5mm) (Hwang et al., 2020). Primary microplastics include items such as

microbeads found in cosmetic products and plastic fibers used in synthetic fabrics. Secondary

microplastics are those plastics that begin their life cycle as macroplastics but, through their

intended use or natural processes, break down and shed microplastics into the surrounding

environment (Hwang et al., 2020). Frequent sources of microplastics include car tires, sneakers,

and single-use plastic packaging. Certain items, such as car tires and sneakers, shed plastics as

part of the normal wear and tear of the item. Car tires lose their tread as the tire makes contact

with the road surface, producing microplastics, similar to sneakers as people walk from place to

place. Car tires are made partially of synthetic rubbers, which are artificial elastomers made

using a type of plastic that is formed from petroleum by-products (Kole et al., 2017). Other

items, such as the vast majority of single-use plastic packaging, generate microplastics after they

are discarded. These plastics make their way into the natural environment and gradually break

down and produce microplastics as the plastic is subjected to winds, UV light, and the abrasion



8

that comes from being picked up in a river or stream and carried downstream (see Figure 1)

(Song et al., 2017).

Figure 1: Diagram of Microplastics Transport Pathways in the Ocean (Ashrafy et al., 2023)

Microplastics can be broken into several size and shape categories, which ultimately

affect where these plastics end up and how they impact the environment. The most common

shapes for microplastics in water and sediment samples are fibers, fragments, beads, films, and

foam (Kooi and Koelmans, 2019). Shape is an important factor in how microplastics interact

with the environment. One study found that shape plays an important role in bioadhesion,

adsorption, and biofouling; fragments adhered to plants greater than films, while films had the

greatest adsorption capacity and development of biofilm (Rozman et al., 2023). Biofouling is the

“accumulation of organisms on submerged surfaces and affects the hydrophobicity and buoyancy

of plastic” (Kooi et al., 2017). As microplastics continue to break down once they enter the

environment, their shapes and sizes are constantly changing. In freshwater, the density of the

plastics plays an important role in transport and where they are concentrated; more dense

microplastics tend to sink while less dense microplastics often float on the surface of the water



9

(Shamskhany et al., 2021). Density and size influence the rate of sedimentation of plastics,

degradation, and biofouling (Kye et al., 2023). In seawater, a large fraction of microplastics are

fragments and the collection of biofilm on these microplastics with larger surface areas may

affect the sinking rate and thus distribution of the plastics in the water system (Kye et al., 2023).

Transportation through marine environments is highly dependent on particle size, as they

accumulate in different marine compartments based on hydrodynamic parameters (Shamskhany

et al., 2021). In addition, depending on their size, microplastics interact differently with

organisms (Kye et al., 2023). For example, larger microfibers led to a greater immune response

in salmonid and a greater lethality of infectious hematopoietic necrosis virus than smaller

microplastics (Seeley et al., 2023). Microplastics smaller than 20 µm can pass through cell

membranes and those smaller than 1.5 µm can damage various types of cells, such as red blood

cells, and lead to inflammation (Kye et al., 2023; Hwang et al., 2020). These microplastics pose a

great risk to organismal health because of their ability to cross the cell membrane.

Once in the natural environment, microplastics can have a host of negative impacts on the

surrounding ecosystem. While the impacts of microplastics are not yet fully understood, one

concerning trait of microplastics is their ability to interact with harmful toxins in the water

column and facilitate the processes of bioaccumulation and biomagnification (Peng, 2017;

Turner and Holmes, 2015). Due to certain chemical properties of microplastics, chemicals and

other contaminants can adhere to the surface of microplastics, allowing these chemicals to reach

a higher concentration on the plastic than otherwise found in the ambient environment (Rochman

et al., 2013). When the microplastic is consumed by an organism, these higher concentrations of

contaminants are transferred into the organism. If this organism consumes high volumes of

plastics, the contaminants found in the organism can reach levels far beyond what would be



10

normally found in the absence of plastics (Ribeiro et al., 2019; Ding et al., 2022). Notably, these

contaminants can be transferred between trophic levels, potentially increasing risk of

bioaccumulation (Carbery et al., 2018). The amount that microplastics actually adsorb is highly

variable, as it is based on environmental factors ranging from salinity to hydrophobicity, making

the risks of increased toxicity and bioaccumulation different for each environmental system

(Rafa et al., 2024).

In addition to the effects of bioaccumulation, some studies have attributed the

consumption of microplastics to a variety of negative health effects on the organisms that

consume them. They may invade animals’ tissues and organs, causing diminished feeding,

growth inhibition, reduced mobility, and other conditions (Wang, 2019). Microplastic

consumption has also been shown to impact the antioxidant abilities of fish, leading to

neuromuscular damage (Elizalde-Velázquez, 2021). Animals ingesting microplastics may also be

ingesting toxins that accumulate on the microplastics. For example, chemicals in microplastic

fibers can act as endocrine disruptors, which skew fish and oyster sex ratios (Kim et al., 2015).

In oysters, polyethylene terephthalate (PET) plastic may initially reduce growth for up to three

months but do not seem to significantly affect overall growth (Sorini et al., 2021). However, the

ratio of males to females was much lower than in control populations, being 54%-44% instead of

82%-18%, respectively, which could affect conclusions about oyster size (Sorini et al., 2021).

The effect of microplastics on human health is a particularly novel research focus, but the

results of preliminary studies are an indication of the importance of continuing to expand this

area of research. Microplastics have recently been found in multiple tissues in the human body,

including human livers, kidneys, placentas (From Pollution to Solution, 2021). Plastics

themselves may be harmful, and this is continuing to be understood, but microplastics are



11

particularly threatening because of the other chemicals that they contain or absorb. The

International Union for the Conservation of Nature found that chemicals found in plastic

products are carcinogenic and may cause developmental, neurological, reproductive, and

immune disorders (Issues Brief: Marine Plastic Pollution, 2021). These new findings undermine

the urgency with which new research about the presence of microplastics and their impact on our

communities must be initiated.

As microplastic research moves to the forefront of environmental toxicology, it is

important to make collection and processing methods consistent across studies. Currently, nets,

pumps, sieves, bottles, and buckets all provide a potential source for variation leading to different

counts of microplastics across studies (Prata et al., 2019). While the type of sample collected

(surface water or sediment) plays a large part in the collection methodology, there are still factors

that can be standardized to avoid counting biases in studies. Once samples are collected,

researchers are tasked with the problem of turning a volume of water or sediment into a slide that

can be quantified. Acidic, alkaline, and enzymatic solutions are used to digest samples and each

affects the microplastics in the sample, such as through melting, yellowing, or degradation (Prata

et al., 2019). Additionally, depending on the purity of the solution used and the lab environment,

microplastics can be added to samples during processing.

Developing a consistent methodology based on previous research is crucial to make

informed decisions about plastic regulation. Once a standardized methodology is in place,

researchers can correlate population density and land use to microplastic concentrations (Yonkos

et al., 2014). Additionally, variation exists between the size of particles in different bodies of

water. Bikker et al. (2020) found that the mean particle concentration of the Chesapeake Bay was

0.160 particles per cubic meter. The size range for this study was 0.007 particles per cubic meter



12

and 1.245 particles per cubic meter. The Chesapeake watershed has also been analyzed for

microplastics, namely in four tributaries of the bay which were found to have microplastic levels

between 5 to 200 g/km2 (Yonkos et al., 2014). Davey et al.'s (2023) microplastic research within

the Anacostia watershed identified sites along Nash Run and found levels of 24-127

microplastic particles per liter (counted using a dissecting microscope). The presence of

microplastics in local waterways are relevant issues for local agencies such as the Anacostia

Riverkeeper who contribute to citizen science to define the nature of plastic pollution in the

region (O’Donnell, 2019). With consistent processing methods, these concentrations could be

compared more confidently.

One of the most pressing issues in the realm of microplastics research is the handling of

contamination and error (Bogdanowicz et al., 2021). The three major error types we will discuss

are false-positives, false-negatives, and exclusion bias. We will first examine a typical

microplastic processing protocol intended to prepare slides for fluorescence microscopy. The

major steps begin with sample digestion to remove organic matter from the sample which could

clog future filtration steps. As discussed previously, this step can either be too weak, failing to

sufficiently remove organic material, or too strong, potentially altering the integrity of the

microplastic particle in a melting-like process (Prata et al., 2019). This step is followed by

filtration through a glass-fiber filter, with other steps added to adjust to specific sample

properties. For example, density separation may be added to samples with excesses of sediment

present (Nava & Leoni, 2021). Next, a fluorescent stain is applied to the samples which allows

for the dye to adhere to plastics and fluorescence under specific wavelengths (Sturm et al., 2023).

Finally, the slide is examined using a fluorescence microscope and manual counting.



13

The utility of fluorescence microscopy is that a fluorescent stain can be added to a liquid

sample which sorbs primarily to plastic particles within solution. In our case, Nile Red, a large

hydrophobic organic molecule, attaches to plastic particles in the solution. The stain can be

excited by various wavelengths of light and exhibits fluorescence which can then be

distinguished from the background more easily than with light microscopy (Shruti et al., 2022).

Throughout this process there are several opportunities for exclusion, including sorption

to the sides of a sampling bottle, never allowing the samples to be present on the filter. Another

source of error is the overestimation of microplastic counts as a result of naturally fluorescing

materials which resemble fluorescent plastics, such as undigested organic material (Sturm et al.,

2023). However, the primary form of error that concerns us is the overcounting of plastics due to

in-lab contamination before the application of the fluorescent dye onto the slides. Some

traditional methods for avoiding this contamination include covering samples in aluminum foil

when not being worked with directly, avoiding plastic-laden clothing, and air filtration systems

(Bogdanowicz et al., 2021).

Many of these issues can be mitigated by using more analytical techniques such as

Pyrolysis-Gas Chromatography-Mass Spectrometry which measures the vaporized plastics

samples for mass (Seeley & Lynch, 2023). While this technique could improve accuracy and

standardization, such instruments are extremely expensive and inaccessible for many

laboratories, and especially for citizen scientists.

This study seeks to develop a replicable and cost-effective method of microplastic

quantification in riverine water samples. Specifically, this study aims to minimize the risk of

cross-contamination during the digestion and isolation phases while maintaining the integrity of

the plastics contained within the sample. To accomplish this, collected riverine water samples



14

were pre-stained with Nile Red prior to any digestion or processing to minimize the risk of

ambient microplastic contamination from being quantified in the final counts. This procedure

was then applied to a study of the impacts of a large urban area on the concentration of

microplastics in its surrounding waterways.

Sampling

Site Selection

The initial goal of this project was to quantify the University of Maryland’s contribution

to the prevalence of microplastics in the surrounding waterways, specifically the Paint Branch

Creek watershed. As a result, we chose to sample from sites above, below, and on campus (see

Figure 2). These sites were primarily chosen based on their relative location to campus and

secondly based on sampling accessibility.



15

Figure 2: Anacostia watershed study site locations

The choice to focus on Paint Branch Creek was a practical choice so that we could

consistently collect samples from the sample locations and compare them among sites. This is a

more specific study region than the originally proposed larger Chesapeake and Anacostia

watershed.

Sites for this study were chosen to capture both above and below the University of

Maryland campus while staying on Paint Branch Creek. This sample range was a scaled down

version of initial plans for study of the Anacostia watershed and greater Chesapeake Bay. Our

final sample list was used to look more specifically at the microplastics impact of the University



16

of Maryland College Park campus impact. Our final sites were each nearby a landmark, one of

the seven following: Bladensburg Waterfront Park, the Northeast Branch Anacostia River Stream

Gauge, the College Park Airport, the Brendan Iribe Center for Computer Science and

Engineering, Lot 11 at the University of Maryland, the Rita's Italian Ice of College Park, and

Paint Branch Golf Complex. These landmarks were used to refer to the locations of our sites

throughout the collection and processing of samples.

Table 1: Site Names, Numbers, and Coordinates

Site Identifier Site Number Coordinates

Bladensburg Waterfront Park 1 38°56'11.4"N 76°56'28.0"W

NE Gauge 2 38°57'36.4"N 76°55'32.9"W

Airport 3 38°58'38.5"N 76°55'06.5"W

Iribe 4 38°59'22.2"N 76°56'04.0"W

Lot 11 5 38°59'33.9"N 76°56'07.1"W

Rita’s College Park 6 38°59'33.9"N 76°56'07.1"W

Paint Branch Golf Complex 7 39°00'00.9"N 76°55'57.2"W

This initial site list additionally included a river bank directly west of the College Park

Ikea for the purpose of capturing parking lot runoff. However, this site was not readily available

for sampling due to fencing and forest brush, and was cut from our sample list. Furthermore, we

chose to collect surface water samples from the shoreline due to ease of access. Surface water

refers to water at the top of the water column.



17

An important consideration for sampling results from our ultimate site list is that rather

than open waters, we chose to collect from creek surface waters, which presents issues with leaf

litter and organic material in our samples.

Site Description

Bladensburg Waterfront Park is a public park in Prince George’s County with a history of

pollution and recent efforts toward remediation. Notably, Bladensburg Waterfront Park (see

Figures 3-6) is our only tidally influenced site due to its size and proximity to more open water.

Bladensburg is the site with the most visible debris upon selection for our study (see Figure 6).

Figure 3: Bladensburg Waterfront Park Sampling Site



18

Figure 4: Bladensburg Waterfront Park Sampling Site

Figure 5: Bladensburg Waterfront Park Sampling Site



19

Figure 6: Bladensburg Waterfront Park Sampling Site

The Northeast Branch Anacostia River Stream Gauge (see Figures 7-9) is located north

of the Bladensburg sampling site across the street from an apartment complex and near an urban

park. This location was chosen as measurements of water speed can be measured using data

collected by the stream gauge (see Figure 8). Notably, this site involves two portions of the river

re-mixing after separation at the bridge (see Figure 7); yet before a large rapid shown in Figure 9.



20

Figure 7: Northeast Branch Anacostia River Stream Gauge Sampling Site

Figure 8: Northeast Branch Anacostia River Stream Gauge Sampling Site

Figure 9: Northeast Branch Anacostia River Stream Gauge Sampling Site

The College Park Airport sampling site (see Figures 10-12) is slightly north of the

Northeast Branch Anacostia River Stream Gauge, directly adjacent to the College Park Airport



21

runway. Similarly, there is a nearby urban park by this site as well as an outflow pipe (see Figure

13).

Figure 10: College Park Airport Sampling Site

Figure 11: College Park Airport Sampling Site



22

Figure 12: College Park Airport Sampling Site

The Iribe sampling site (see Figures 13-15) marks our first on-campus site, directly across

from the Iribe Computer Science building on the University of Maryland campus. This site is in

a forested area, surrounded on both sides with tree buffers. This site is directly below a fairly

busy road in the city of College Park, US Route 1, as shown by the bridge in Figure 13.



23

Figure 13: Iribe Sampling Site

Figure 14: Iribe Sampling Site

Figure 15: Iribe Sampling Site

The Lot 11 sampling site (see Figures 16-18) is similar to the Iribe site with a walking

bridge above the stream. There are a series of sandy beaches along with banks of the creek in this

region as well as the merging of two sources, as shown by the adjoining stream shown in Figure

16.



24

Figure 16: Lot 11 Sampling Site

Figure 17: Lot 11 Sampling Site

Figure 18: Lot 11 Sampling Site



25

The Rita’s sampling site (see Figures 19-21) is below an urban bridge as shown in Figure

19. The banks of the river are surrounded by a rocky, sandy beach and forested area on one side

and a human-built wooden fence on the other side (see Figure 20). This portion of the creek is

directly adjacent to a parking lot and urban park. Over the course of this study, construction

began on the road directly next to this site, leading to the creek.

Figure 19: Rita’s Sampling Site



26

Figure 20: Rita’s Sampling Site

Figure 21: Rita’s Sampling Site

The Golf Course sampling site (see Figures 22-24) is the most remote site we sampled at,

with the nearest landmarks being the paved Paint Branch trails. There is a walking bridge above

this site and a golf course within a quarter mile of our sampling site.



27

Figure 22: Golf Course Sampling Site

Figure 23: Golf Course Sampling Site



28

Figure 24: Golf Course Sampling Site

Collection

We collected two one liter samples from each site on a monthly basis along with one 1L

blank sample using pre-collected distilled water from a laboratory setting. The purpose of taking

duplicate samples is to be able to compare within replicates. However, only two samples were

taken at each location due to material and transportation constraints.

Our intention for our collection process was to have 1 L duplicate samples as well as 1 L

of a control for each sampling location and collection time. To minimize the time required to

process each batch of samples, we opted for one control bottle per run filled with 1 L of filtered

deionized water, opened each time a sample was collected. Originally, we attempted to collect

our samples by placing a 1 L glass bottle directly into the running stream and allowing it to fill



29

with water. However, this created issues with air pocketing in which air leaving the bottle

displaced sediment within the water column. This displacement biased samples due to increasing

the collected sediment content within our samples. Due to the heavier microplastics which

disproportionately exist in the sediment, this sampling method needed to be rethought.

Furthermore, we wanted to collect within the middle of the water column, not the surface of the

stream.

Our proposed solution was the equivalent of a two-sided jar, which would allow for

stream water to flow through the apparatus before being closed on both ends simultaneously.

This would allow for an effective “snapshot” of the stream at a given moment in time. When

placed parallel to the flow of water, the apparatus can easily allow water to flow through and

settle, before rapidly and simultaneously closing from both sides of the tube to collect 1 L of

water. On site, we immediately funneled this into a 1 L glass bottle. Our apparatus used PVC

piping fitted to the specific dimensions to prevent disturbance, as illustrated by Figure 25.



30

Figure 25: Technical Diagram of Sampling Apparatus Showing Dimensions



31

Table 2: Sampling Apparatus Parts List
Tag Part Number
1 2.5" ASTM D1785 PVC Pipe

2 Inline Tee Reducing Adapter, Female Socket Connector
(2.5" to 2")

3 Straight Bushing Reducing Adapters with Collar, Male Socket Connect × Female
Threaded Pipe
(2" to 0.75" with NPT Threads)

4 Straight Adapters, Female Socket Connect × Male Threaded Pipe
(0.75" with NPT Threads)

5,7 0.75" ASTM D1785 PVC Pipe

6 0.75” Tee Female Socket Connect

8 2.5" ID Shaft Mounted Seals



32

Figure 26: CAD Representation of Sampling Apparatus

Notably, filtered water samples were run through the apparatus and tested as a blank to

ensure that the apparatus itself was not shedding microplastics into our samples. While a

non-plastic apparatus would have been preferred to remove any concerns of contamination,

budget constraints and the desire to develop a replicable and cost-effective methodology for use

by citizen scientists prompted us to stick with the PVC sampling apparatus, given the confidence

that it was not adding excessive contamination to our samples based on the blanks. To collect a

sample, we lowered this apparatus below the surface of the water and allowed for the stream to

flow through for 15 seconds before pulling a cord closing the apparatus, trapping 1.175 L of

streamwater inside. Next, the apparatus was pulled out of the water and manually opened on one

side allowing for water to pour into a 1 L glass bottle, assisted with a funnel to ensure sample



33

retention. Samples were shaken to ensure even distribution of the contents and leveled to exactly

1 L of water. Samples were then labeled based on the sampling run month, site, and replicate.

Figure 27: Apparatus Being Used to Collect a 1 L Sample



34

As samples were collected at each site, weather patterns as well as research observations

were recorded. Importantly, recent rain events were noted along with the approximate height of

the stream, turbidity, and qualitative flow rate (noted as fast or slow based on visual inspection).

We organized our samples with a three number code written on each bottle. The first

number indicates which collection run the sample is from, the second the sampling location, and

the third which replicate that sample was. Numbers ending in 00 indicate a control sample. This

system was used to help more effectively organize and keep track of our samples.

Table 3: Collection Dates and their Numerical Code
Collection Run Collection Date

0 Summer 2022 (May 4, July 6, and July 8 - depending on sample)
1 Oct. 16, 2022
2 Nov. 20, 2022
3 Jan. 7, 2023
4 Feb. 18, 2023
5 March 11, 2023
6 April 27, 2023
7 June 3, 2023
8 June 24, 2023
9 July 8, 2023
10 Aug. 26 and 27, 2023

Processing

Determining an optimal processing method for our samples was the most time-intensive

part of the project. We used our collection method to collect extra samples as needed to test the

effectiveness of each part of the process as we developed it. The full collection and processing

protocol is available below (see Appendix A).



35

Acidification and Digestion

Our initial attempt was to directly pour each sample over a filter to then quantify the

remaining microplastics; however, sediment and organic matter clogged the filters necessitating

further filtration. These were glass fiber filters without binder resin and a pore size of 0.7

microns (MillporeSigma Glass Fiber Filters).

Figure 28: Filtration Set-Up

https://www.fishersci.com/shop/products/glass-fiber-filters-without-binder-3/AP4004705


36

Figure 29: Sediment-Clogged Slides Which did not Receive Density Separation

For the first few sampling runs, we acidified each sample with 1 mL of nitric acid to

prevent the growth of organic material during storage. We then digested each sample at a later

date. Each sample was digested with 85% purity potassium hydroxide pellets sourced from

Thermo Fisher Scientific Chemical, (added at 1M concentration) and then placed in a heated

shaker for 48 hours. This process could lead to spilling and explosions so the bottles were

required to be loosely capped to avoid pressure build. In later months, we were able to digest

directly after collection and therefore skipped the acidification step.

In troubleshooting this digestion process, we found that loosely capping our bottles led to

a pressure build-up within the media bottle, especially in samples with high amounts of sediment

and possible organic matter. As a result, these pressurized bottles were leaking sample which had

essentially spilled through the cap. For future tests, we ensured that the bottles remained tightly

capped to prevent this problem.

Additionally, we considered a couple different digestion methods, including hydrogen

peroxide (Tan et al., 2022). We limited our digesting reagent option to potassium hydroxide and

hydrogen peroxide because we wanted to avoid overly harsh digestions which may degrade the



37

plastics within the sample. Hydrogen peroxide is more often used for plant and sediment-rich

water samples and has been found to be less harsh on certain types of plastics including

polyethylene terephthalate (PET) (Pfeiffer & Fischer, 2020). However, we opted for the

potassium hydroxide digestion method due to the familiarity of this method in our lab and

previous success with an existing methodology which we modified.

Staining

We chose to stain our samples with Nile Red stain to make the microplastics fluoresce

when viewed through the microscope during the counting process. Originally, we tried the

traditional method of adding Nile Red after the sample had been put onto a filter. However, as we

added more steps into our processing procedure, we opted to add the stain to the sample before it

was put onto the filter. This would prevent any microplastics that contaminated the filter through

in-lab contamination from being stained and fluorescing later.

We found no noticeable difference in the effectiveness of Nile Red stain at different

concentrations in the 1 L sample. We tried adding 0.5 mL, 1 mL, 2 mL, and 5 mL of Nile Red to

samples, and the microplastics were just as fluorescent. Similarly, we added higher volumes of

Nile Red dye to the samples by testing 10 mL and 15 mL additions. None of the tested samples

had noticeably different levels of fluorescence. As a result, we chose the 5 mL volume in order to

reasonably conserve materials while having a slight excess volume of Nile Red available in case

a sample had a particularly high concentration of microplastics.

Allowing for Nile Red to sit for several hours without analyzing samples can result in

Nile Red aggregates, similar to micelles, due to the hydrophobic nature of the dye (Mino et al.,

2023). This could potentially cause concern given that we prepared Nile Red and allowed it to sit

for up to a month before use as well as adding Nile Red to our digested samples and allowing for



38

these samples to sit for days or weeks. However, this methodology involved sending the samples

through a sieve and only collecting the particles which are greater than 20 µm in diameter. These

Nile Red aggregates are known to be smaller in size, estimated to be closer to 130 nm, and are

therefore not a concern with the proposed method (Sutter et al., 2007).

We tried several different iterations of this technique including staining after digestion,

before digestion, and simultaneous to digestion. We found no notable difference between these

methods and therefore chose to stain simultaneously with digestion for the purpose of

streamlining our laboratory processing and decreasing handling of our samples. As a result,

immediately following the addition of 56 g of potassium hydroxide pellets, 5 mL of Nile Red

was added to the sample media bottle. Each sample was left in the shaker for 48 hours after this

step to ensure mixing and distribution of the stain throughout the sample.

Density Separation

Another issue we found - specifically from working with surface water samples - was that

the samples quickly clogged our glass-fiber filters creating layers such that the samples could not

be observed effectively under the microscope.

Our initial method for solving this involved decanting off the top portion of the digested

sample, accounting for approximately 800 mL of the 1 L sample. Following this step, the

remaining sediment-rich sample was divided into 250 mL round-bottom centrifuge tubes, which

were then balanced with filtered water and centrifuged at 2500 rpm for 10 min. The top portion

of water was then poured through a baked glass-fiber filter. This second decantation was done

carefully to avoid clogging the filter. Next, the remaining sample was washed with additional

water to dilute the potassium hydroxide present in the sample. This washing step is important to

reduce the precipitate formed from future zinc chloride and potassium hydroxide interactions.



39

The washed sample was centrifuged again at 2500 rpm for 10 minutes and decanted into the

filtration set up. Next, approximately 100 mL of 11 M (1.5 g/mL) zinc chloride (Zinc chloride,

98%, extra pure, Thermo Scientific Chemicals) was added to the centrifuge tube and the sample

was shaken. This sample was centrifuged at 2500 rpm for 10 min and decanted into the filtration

set up. This process retained a couple of the same issues that we were having before in that the

filters continued to clog and therefore multiple filters had to be used per sample to get accurate

counts. In addition, a huge amount of materials was being used to process each sample. For the

purpose of making this method both accessible and not cost prohibitive, we decided to change

this initial method by modifying the original sample. We additionally tried different separation

methods including gravity separation using a separatory funnel but found that the separation

layers were not sufficiently clear within a 48 hour period to warrant further investigation.

Figure 30: Centrifuge used for these analyses



40

Figure 31: Sample Post-Density Separation

Our altered method begins by running the digested and stained 1 L sample through a 20

µm sieve and rinsing thoroughly with filtered deionized water. Following this rinse, a solution of

filtered 11.0 M (1.5 g/L) zinc chloride was used to rinse the sieve directly into a 50mL centrifuge

tube. This step reduces the need to move between glassware and reduces risk of both

contamination and sample loss. Notably, such an alteration of the method reduced our study to

only consider large microplastics over 20 µm in size. However, the amount of material used per

sample is much smaller and interactions between potassium hydroxide and zinc chloride are

minimized.

The zinc chloride collected sample is then centrifuged at 2500 rpm for 10 min to allow

the sediments to pelletize. The top layer of zinc chloride is then decanted into the filtration set up

to prevent excess sediment from clogging the filter. Then approximately 15 mL of zinc chloride

is added to the remaining sediment compacted at the bottom of the sample. This sample is then

shaken to resuspend the sediments and centrifuged again at 2500 rpm for 10 min to ensure



41

microplastics were not caught in the previous sediment pellet. The layer of zinc chloride is then

decanted through the sample filter. Finally, the sample is washed with approximately 100 mL of

filtered deionized water and then covered with aluminum foil and allowed to dry for

approximately 10 min.

Some important precautions had to be taken for this method included a thorough wash of

the sieve between samples as well as continuously covering the filtration set-up with aluminum

foil when not directly decanting liquid through it in order to minimize in lab ambient

contamination.

Quality Control

We conducted a rigorous stepwise check of our developed methods to ensure that the

developed procedure was accurate and was not introducing unknown sources of error into our

study. To do this, we created plastic fragments and shavings in three separate size categories

(small, medium, large) that were prestained with Nile Red and added a set volume of 5ml to both

control samples and actual unstained environmental samples to examine the impact of our

procedure on these plastics throughout the process. These samples were then passed through our

procedure, with each sample being stopped at a different step in the process and passed through a

filter to examine the recovery rate of the spiked microplastics. We found through this process

that our digestion process was not resulting in any overdigestion or degradation of the plastic

particles nor was any specific step resulting in a large loss of the spiked plastics. This quality

control stepwise process gave us the confidence that our developed method was fairly accurate

and did not result in any undue contamination or microplastic loss, allowing us to begin

collecting and processing field samples.



42

Quantification

Once the samples were processed, we were left with dozens of slides that could be

counted at our leisure; it required that we manually counted what was on each slide. Due to the

nature of what was on each slide, we needed to be strict on what we were classifying as

microplastic, so that each person was observing the same counts.

In order to count the samples, each slide was analyzed with a Leica DM LA microscope

at 40x magnification. We specifically use fluorescent microscopy with excitation at 510 nm.

Starting from one end of the sample, the person counting would systematically scan through the

entire slide, counting what they saw on each frame before moving on to the next. Most used a

snake pattern to traverse the entire slide.

Each sample was counted by two to three people, one time per person, in an attempt to

account for variability in counting. Each sample was counted twice initially, and, if the total

particle count differed by greater than 20%, then it was counted a third time. Additionally, we

separated the counts into categories for the types elements we saw on the slides to help with

consistency. We named the categories the following: pinpoints, small particles, large particles,

fibers, and nodes (see Tables 4 & 5). The categories were classified qualitatively with pinpoints

being dull points of light with ill-defined edges which could not be focused on at the 40x

magnification. Large and small particles were more traditional fragments which could be seen

and easily identified as microplastics but were delineated based on size. The distinction was

based on team consensus from our counting training to get everyone on the same page. Fibers are

long, string-like pieces of plastic. Finally, nodes are nearly perfectly circular fluorescing objects

which often were found in clusters. These nodes differ from other described plastics as typically

plastics we found had more jagged edges whereas nodes were far smoother. The pinpoints



43

category was added after a few initial counts. Before, they were counted under small particles,

but some people would count them while others would not, being unsure if that was a

microplastic or something else. Interestingly, the pinpoints category mostly include fragments

smaller than 20 µm, meaning that they should not have made it through our sieving process. As a

result, pinpoints likely constitute some type of contamination from our processing method,

potentially from a self-fluorescing species. As a result, the pinpoints were counted in all samples

but were not considered as part of our environmental samples.



44

Table 4: Guidelines for Our Classification of Different Items that Fluoresce

Pinpoint Small Large Fiber Node

Description Small, dull
light, not
sharp/defined
edges

Clearly
fluorescing,
sharp defined
edges, small

Clearly
fluorescing,
sharp defined
edges, large,
easily visible
at 10x
magnification

Clearly
fluorescing,
long and thin

Clearly
fluorescing,
very circular,
often
occurring in
groups

Example
Photographs



45

Table 5: Comparison of Fluorescent and Light Microscope View of Different Plastic Shapes and
Types

Microplastic Shape Type of Microscope View

Nile Red Fluorescent
Microscope

Light Microscope

Fragment

Fiber

Node

Comparison - small and large



46

Pinpoint

Because of time limitations, samples from only three sites were counted in addition to a

control for each sample run. The sites were Iribe, the Golf Course, and Bladensburg. From each

of the sites, a sample pair from each season was counted (see Appendix B).

Throughout this process, we looked into several other methods for quantification

including use of Sedgwick rafter cells and full slide counting. Sedgwick Rafter cells are

microscope slides with a precisely printed grid allowing for high resolution counting within a

defined area, and by extension volume, of the slide. We considered incorporating this method

into our counting by taking five representative 1 mL samples from our field-collected samples

and pipetting this subsample onto the Sedwick Rafter cell (Pratesi et al., 2020). This method

would allow for extrapolation from our subsample concentrations to find the total number of

microplastics in the collected sample overall. One benefit of this method is the lack of required

sample handling, which would then decrease risk of lab-introduced contamination. However, we

opted against this method due to the size of our collection bottle. We are already collecting very

small samples of large water bodies. As a result, taking even smaller subsamples contributes to

potentially greater levels of sampling bias. We therefore chose to count all of the plastics on each

processed sample filter, which was prepared as described above. Another issue found with this

method was the presence of sediment in our field-collected samples which obstructed our view



47

of microplastics under the microscope. Such an issue can be circumvented with density

separation, which then defeats the purpose of directly looking at our sample media under the

microscope. Additionally, we found issues with identifying plastics on the Sedwick Rafter cells

with the plastics grouping together as well as migrating to the sides of the glass cell. This is

likely due to our samples being held within an aqueous solution, which repels the more

hydrophobic plastics particles making the Sedgewick Rafter cell method infeasible in our

environmental media.

After ruling out the Sedgewick Rafter method, we chose to pursue full filter counting by

moving around the slide in standard search patterns described above. We trained all members of

our team on the same three slides and counted in repetition until every member of the team was

counting approximately the same number and distribution of microplastics of each type.

Following analysis of these results, we found no counter bias and proceeded to count each of our

selected samples in duplicate. Any samples which showed a discrepancy between counters were

then subjected to a triplicate count and all three measurements were averaged.

Data and Results

Particle Classification Results

The composition of sample particles were categorized into large particles, small particles,

fibers, and nodes (see Table 6). Small particles constituted the greatest percentage of total

particles sampled at 75%, large particles were 18%, fibers were 3%, and node particles were 4%

(see Figure 32).

All samples (N = 28, n = 27) were collected as duplicates and each duplicate was counted

twice. The counts for each duplicate sample were then averaged yielding one average count for

each duplicate. For any counts that were egregiously inconsistent between raters (variation by



48

greater than 20% between counters), we conducted a triplicate count. This triplicate count was

then averaged with the original two counts. One sample was excluded from this study due to an

inability to effectively quantify the sample particles with our procedures.

Table 6: Descriptive Statistics of Average Sample Particle Distribution per Liter across 27
Samples
Particle Size Mean SD Median Min Max Range

Small 842 446 773 269 1873 1604

Large 199 212 132 9 919 910

Fiber 32 21 30 6 81 76

Node 42 66 12 1 231 230

Total 1115 646 1067 346 2879 2533

Figure 32: Average distribution of particle types across samples



49

Seasonality Results

We conducted a seasonality analysis to understand average sample particle quantities

across each of the four seasons. We found that the average particles were highest in Spring at (x̄

= 1603); followed by Fall (x̄ = 1117); then Summer (x̄ = 999); and lastly Winter (x̄ = 724) (see

Figure 33). Within each season the composition of particle types was largely consistent with the

overall averages (see Figure 32). Nodes were an exception and were most prevalent in Spring

and Summer; however, largely absent in the Fall and Winter. Spring had 71% small particles;

20% large particles; 3% fiber particles; and 6% node particles. Fall had 74% small particles; 22%

large particles; 3% fiber particles; and 1% node particles. Summer had 81% small particles; 11%

large particles; 4% fiber particles; and 4% node particles. Lastly, Winter had 81% small particles;

16% large particles; 2% fiber particles; and 1% node particles.

Figure 33: Average Total Particles by Season



50

Location Results

We conducted a location analysis to understand average sample particle quantities across

each of the three sites and our control. We found that the average particles were highest in

Bladensburg at (x̄ = 1558); followed by Control (x̄ = 1027); then by Golf Course (x̄ = 926); and

lastly Iribe (x̄ = 875) (see Figure 34). Within each location the composition of particle types was

largely consistent with the overall averages (see Figure 32). Bladensburg had 75% small

particles; 17% large particles; 2% fiber particles; and 5% node particles. The Control had 67%

small particles; 23% large particles; 4% fiber particles; and 6% node particles. Golf Course had

74% small particles; 20% large particles; 4% fiber particles; and 2% node particles. Lastly, Iribe

had 83% small particles; 12% large particles; 3% fiber particles; and 2% node particles. We then

compared the results of our seasonality and location analysis to understand how the two are

related (see Figure 35).

Figure 34: Average Total Particles by Location



51

Figure 35: Average Particles by Location and Season

Figure 36: Total Microplastics Count Between Sites and Between Seasons. Error bars represent
the range between the totals of two samples at each location (only one control sample was taken
for each date, so there is insufficient data for error bars on control samples).



52

ANOVA Statistics

ANOVA Assumptions

We conducted a series of one-way ANOVAs to compare particle totals across all four

seasons. We then conducted a second series of one-way ANOVAs comparing particle totals

across our three sites and control. Before running these ANOVAs, we conducted the

Shapiro-Wilks test of normality and found our data to be normally distributed following the

removal of one outlier. This outlier was identified and removed from the dataset due to it falling

outside of the normal distribution. There were excessive particles in this sample which could not

be effectively quantified with our procedures. Next, we performed a Levene’s test to assess

homogeneity of variance where all tests were not statistically significant indicating the

assumption was satisfied. All data were assumed independent.

ANOVA Statistics

All analyses were conducted in R Studio (Version 2023.09.1) and Microsoft Excel

(Version 16.0.17324.42305). The ANOVA series addressed the comparison between particle

totals across all four seasons and site locations, respectively. Each series of ANOVAs were run

looking at small particles, large particles, fiber particles, node particles, and total particles. These

categories were created to distinguish the different particle types that were visible on our

samples. The first series compared these particles to seasons and the second to location. Tukey’s

t-tests were performed for any significant result to understand which variables significantly

differed from one another.



53

The first series of one-way ANOVA tests were conducted to see if there were any

significant differences in quantities of particle sizes across four seasons. There were no

significant differences for small particles F(3,23) = 1.997, p > 0.05; large particles F(3,23) =

1.625, p > 0.05; or total particles F(3,23) =2.694, p > 0.05. There were significant differences in

fiber particles F(3,23) = 5.812, p < 0.05 and node particles F(3,23) = 3.954, p < 0.05. Tukey’s

post-hoc tests were run and found a significant difference in quantities of fiber particles between

Spring and Winter (M = -36.43, p < 0.05). Another set of t-tests were run for nodes and found a

significant difference in quantities of nodes between Spring and Winter (M = -93.98, p < 0.05);

and Spring and Fall (M = 86.26, p < 0.05).

Table 7: One-Way Anova Results for Significant Differences in Quantities of Particle Sizes
Across Seasons
Particle Size DFn DFd F p

Small 3 23 1.997 0.14

Large 3 23 1.625 0.21

Fiber 3 23 5.812 0.004*

Node 3 23 3.954 0.021*

Total 3 23 2.694 0.07

*p < 0.05.

A second series of one-way ANOVA tests were conducted to see if there were any

significant differences in quantities of particle sizes across site locations. There were no

significant differences for small particles F(3,23) = 2.425, p > 0.05; large particles F(3,23) =

0.782, p > 0.05; fiber particles F(3,23) = 0.490, p > 0.05; node particles F(3,23) = 2.223, p >



54

0.05; and total particles F(3,23) =2.694, p > 0.05. Tukey’s post-hoc tests were not run since there

were no significant differences found.

Table 8: One-Way Anova Results for Significant Differences in Quantities of Particle Sizes
Across Site Locations
Particle Size DFn DFd F p

Small 3 23 2.425 0.09

Large 3 23 0.782 0.52

Fiber 3 23 0.490 0.69

Node 3 23 2.223 0.11

Total 3 23 2.050 0.14

*p < 0.05.

Interrater Agreeableness

To ensure different counters were not introducing over or undercounting biases into the

results, a statistical analysis of individual counters was done. Each count from all counters was

assigned a percentage differential compared to the other counting replicates of that sample. The

percentage was calculated by finding the average count for the sample and dividing the

difference of the individual count to the average by that same overall average. A percentage was

chosen over a raw difference to avoid the true microplastic count influencing any potential bias.

The mean and standard deviation of each counter’s percentage differentials was calculated and

used to administer a z-test (known population variance) with the null hypothesis that their mean

differential was zero.



55

Table 9: Two-Way Z-Test for Counter Differentials Significantly Different from Zero
Counter Number n 𝜇 σ z p

Counter 1 6 0.10800 0.251 0.430 0.834

Counter 2 6 0.00539 0.222 0.024 0.999

Counter 3 6 0.07380 0.125 0.590 0.778

Counter 4 7 -0.0354 0.170 0.208 0.918

Counter 5 8 -0.0601 0.220 0.273 0.893

Counter 6 7 -0.0669 0.105 0.637 0.762

Counter 7 8 -0.0211 0.121 0.174 0.931

Counter 8 6 0.05003 0.086 0.582 0.780

Based on this test, we cannot reject that the mean differential for each counter was equal

to zero, zmax = 0.637, pmin = 0.762.

Discussion

Methods Utility and Sources of Contamination

This study aimed to develop a methodology that would be economical, easily accessible,

and representative of popular techniques from the existing literature. Our collection protocol

allowed for significant cost savings, as we used a plastic apparatus with a 1 L capacity that was

able to be built by members of the research team without contracting an external makerspace.



56

One serious limitation of our study is the extensive contamination found in our control

samples, ranging up to thousands of plastic particles in what should have effectively been

distilled water. In order to address this issue, we ran a series of water samples through our

staining, sieving, and density separation process which yielded very small levels of plastics

contamination on our slides, which was more in-line with what we expected from our control

samples. Similarly, we ran pre-filtered water (via Elga LabWater Purelab Water Purification

System) through our sampling apparatus and also found minimal contamination. These values

were nowhere near our actual control samples, suggesting that some element of our process was

not represented within this lab test.

Table 10: Microplastics Counts from Lab Control Samples
Condition Microplastics Count (MP/L)
Pre-filtered water with no ambient exposure 15
Pre-filtered water with 45 min ambient
exposure

49

Unfiltered DI water with no ambient exposure 65 ± 22
Unfiltered DI water with 45 min ambient
exposure

130. 5 ± 30. 5

Pre-filtered water exposed to interior of
collection apparatus

83 ± 24

These results suggest that there is minimal contamination that results from our sample

collection. However, because we stain at the beginning of our process, we made the assumption

that contamination which occurred after this initial stain would not fluoresce unless it was

introduced by some self-fluorescing substance (e.g., a mineral of some kind). In an effort to

address these self-fluorescing substances, we added an additional density separation step with

zinc chloride and a nitric acid wash to remove the self-fluorescing substances.



57

As a result we believe that this contamination of our controls occurred before or at the

time of the initial staining. After generating a stained stock solution of potassium hydroxide

(Potassium hydroxide, pellets, 85%, Thermo Fisher Scientific Chemical), we found extensive

plastics present. These plastics were found extensively in all of our described size categories

(including large fragments and fibers). Notably, we added our potassium hydroxide digesting

reagent in pellet form at the same time as our Nile Red stain to our samples. This allows us to be

reasonably confident that the potassium hydroxide pellets represent the majority of our

contamination in both our control and collected samples (see Figure 37). This unfortunate

finding came to light after all sample processing. As a result, we can best address this form of

error by regarding the control level of plastics as background noise, which should be reasonably

similar to those introduced to the collected samples. Therefore, we are still able to look at our

samples comparatively but without a precise count of the plastics content of any individual

sample. In addition, we would make the recommendation that any future microplastics work

limit the use of reagent pellets as they may include plastics within the reagent. The zinc chloride

reagent used (Zinc chloride, 98+%, extra pure, Thermo Scientific) was also analyzed and stained

under the microscope, with no evidence of major contamination.



58

Figure 37: Microscope Photograph of Stained Potassium Hydroxide Solution

One unexplained type of contamination is the spherical node pieces which either

self-fluoresce or are stained by Nile Red. Initially we believed them to be mold of some kind due

to smooth exteriors which are not reminiscent of typical microplastics. Following the discovery

of the potassium hydroxide contamination issue, we propose that the spherical pieces are

potentially seed particles used in the potassium hydroxide manufacturing process. These particles

allow for the salt to pelletize around the plastics which would explain the node category of

plastics. Similarly, we found different levels of background contamination between sample

processing batches. We propose that this batch effect could be attributed to using a different

bottle of potassium hydroxide pellets. As a result, the switch between potassium hydroxide

sources could account for the difference in background noise.

Another protocol that proved difficult was the definitions used during quantification of

microplastics. The categories based on size lended themselves to subjectivity, which was seen

mostly in the variance of counts for the Small category. Although the sieve stage was meant to

remove particles below 20 µm, fluorescent particles smaller than this were visible. While these



59

smaller bits could be representative of larger plastics that broke down during processing, we

counted them separately due to their unknown origin. These particles were counted as pinpoints

to distinguish them from our other classifications which ultimately helped limit counter bias. We

recommend for future fluorescent microscopy studies to use shape over size (fibers, fragments,

beads/round, film), as these distinctions are more easily made given the limits of this

methodology (Davey et al., 2023; Sturm et al., 2023).

While this study did not rely on polymer identification and chemical analysis methods to

classify microplastics, peer studies using both these methods and fluorescent spectroscopy can

shed light on possible origins of these particles. The pinpoints observed in the samples were

likely related to the Nile Red stain. If excess Nile Red molecules were present on the wet filter

following density separation, then they would likely form hydrophobic aggregates, similar to

micelle formation, as a way to reduce interactions with ionic salt or polar water molecules. As a

result, despite expecting Nile Red aggregate radii to be smaller than that of our sieve, the

aggregates could form following density separation and remain on the filter if this formation took

place later (Ray et al., 2019). These aggregates would be small enough that they could appear as

highly fluorescent bits of light and may have been counted within our pinpoints category. As a

result, the pinpoints category is considered as not reliably plastic and is excluded from our

plastics analysis.

A similar analysis was applied to the sample identified and removed as an outlier for

being many standard deviations outside of the population mean (x̄ = 3980.5). Comparison of this

imagery to that of Tong et al. 2021 suggests these many particles were of plastic origin. They

were clearly fragments, resembling the image of polyurethane (PU) particles treated with

potassium hydroxide and stained with Rhodamine B (Tong et al., 2021). This influx in the



60

sample from Summer 2022 at Iribe could potentially be due to a sealant used for nearby

construction. The polymer images also suggest that polypropylene (PP) and polystyrene (PS)

were prevalent in our samples, and that some particles identified as nodes share a resemblance

with those identified as PVC (polyvinyl chloride) by Tong et al., 2021. Polypropylene (PP) is one

polymer commonly associated with tire wear and tear (Kole et al., 2017), and PVC plastics are

commonly used in the pelletization process (Zaman et al., 2016).

Figure 38: Collection of Microscope-view Photographs of Different Types of Plastics. Rows
represent differing processing procedures, and (e) 24 hr treatment with 10% KOH is most
comparable (Tong et al., 2021)



61

Comparable Local Study

Various studies of microplastic contamination have been completed across the

Chesapeake region and its related watersheds with the most similar area evaluated being Nash

Run: an Anacostia/Potomac River tributary about 3 km south of the Bladensburg sampling site

(Davey et al., 2023). This study analyzed both water and sediment samples from Spring 2019 to

Spring 2020. However, statistical significance was only found for their sediment samples. Davey

et al. (2023) found water sample microplastic concentrations of around 10% of the average

values found in this study. This is assuming classification size ranging from > 4000 micrometers

to < 250 micrometers, including categorization by shape for fragments and filaments. These

findings further provide evidence that sources of contamination existed in our samples. Davey et

al. (2023) observed higher microplastic concentrations in Fall season samples; however, they did

not find any significant correlations related to water velocity or sampling parameters (e.g., water

velocity) which was in alignment with our findings. Additionally, chemical composition analysis

and the use of dissection microscopy of samples showed the primary chemical composition of

plastics were polypropylene and polyethylene (Davey et al., 2023).

Effect of Environmental Conditions

As with many waterborne pollutants, microplastic concentrations can be altered by

countless environmental conditions obscuring anthropogenic trends and making them difficult to

distinguish from background environmental variation. These environmental variables can take

many forms, ranging from rain events and weather conditions to tidal fluctuations at sample

locations, both of which can impact base concentration and dispersion throughout the water



62

column. As such, this variation must be considered prior to any conclusions being made about

anthropogenic influences.

When examining water samples within freshwater rivers and streams, it is important to

consider the flow characteristics of the water body at the time of sampling, particularly when

discussing the transport of suspended particles such as microplastics from upstream sources.

Additionally, runoff from rain events can result in increased transport of microplastics through

storm drains, resulting in heightened concentrations in the timeframe immediately following the

rain event. Three water quality variables were analyzed to examine their potential impact of

variation within samples. These variables were turbidity, average discharge volume, and average

water temperature, accessed from the USGS Northeast Regional Stream Gauge (see Figures

39-41). Precipitation and stream gauge height at Northeast Branch USGS Gauge and

Bladensburg Waterfront Park were also analyzed (see Figure 42) (National Centers for

Environmental Information [NCEI], 2024; United States Geological Survey [USGS], 2024;

National Ocean Service [NOS], 2024). These factors differ greatly across seasons, suggesting

that the correlations between these factors across seasons may add noise to the variation

observed in this study. This then prompts caution when drawing conclusions about

anthropogenic influence on microplastic concentration seen in this study. Given these potential

sources of variation, future studies should exercise caution when selecting sampling dates and

locations - accounting for environmental factors - to select dates that will best allow for uniform

replication.



63

Figure 39: Turbidity on the Days that We Sampled as Reported by the Northeast Regional
Stream Gauge (USGS, 2024)

Figure 40: Discharge Volume on the Days that We Sampled as Reported by the Northeast
Regional Stream Gauge (USGS, 2024)



64

Figure 41: Water temperature on the days that we sampled as reported by the Northeast Regional
Stream Gauge (USGS, 2024)

Figure 42: Water Height at both Bladensburg and Northeast Regional Stream Gauge Sites on
Sample Days and Precipitation as Reported at NOAA HYATTSVILLE 0.7 N, MD US station
(NCEI, 2024)

Our sample location at Bladensburg Waterfront Park is in a tidally influenced zone of the

Anacostia River. As such, the daily tidal cycle can influence the concentration of microplastics at



65

this sample location, introducing a potential source of variation in our sample counts. At high

tide, microplastics washed into the Anacostia by the upstream watershed meet at Bladensburg

and likely stagnate at peak high tide, resulting in higher than average concentrations of

microplastics (Ward, 1985). At the same time, microplastics from downstream can be carried

upstream with the incoming tide, and the rising tide can also resuspend plastics that had been left

on-shore by the previous tidal cycle, further increasing concentrations at Bladensburg (Stead et

al., 2020). As such, collections and analysis conducted at a rising or high tide will likely

experience higher levels of microplastics than those collected at a falling or low tide (Feng et al.,

2023). Three out of our four samples collected at Bladensburg were collected during a rising or

high tide, and therefore these samples likely represent an overestimate of the general

microplastic levels. However, these tidal effects can also be confounded by other environmental

variables, such as recent rain events and stream flow at the time of sampling, which could negate

or reduce the effects of tidal trapping.

Table 11: Tidal Positions From Our Sampling Days as Reported by NOAA Anacostia Watershed
Monitoring Site (NOS, 2024)
Date Tidal Position

Fall (10/16/2022) High

Winter (1/7/2023) Low

Spring (4/27/2023) High

Summer (07/08/2023) High

Based on the statistical analysis and trends in the environmental factors, anthropogenic

factors are likely responsible for variation in the results, as the differences in the data are not

https://www.zotero.org/google-docs/?Xd2VNM
https://www.zotero.org/google-docs/?dyz9ge
https://www.zotero.org/google-docs/?dyz9ge
https://www.zotero.org/google-docs/?sZ3OBr
https://www.zotero.org/google-docs/?sZ3OBr


66

fully explained by these trends (see Figure 42). Samples taken during the spring season have the

highest amount of microplastics, however, almost all factors peak in summer (see Figures

39-41). Such trends are consistent with other studies’ findings, such as in Australia where the

abundance of microplastics increased 40 fold after storms (Kye et al., 2023; Hitchcock, 2020).

Conversely, there was a strong correlation between water level and microplastics concentration

in Netravati River, India (Kye et al., 2023; Amrutha and Warrier, 2020). As a result, it is

important to consider both the normal environmental changes such as tidal shifts and

precipitation, which may be biasing our microplastics counts, when addressing potential sources

of anthropogenic introduction. Our Bladensburg site had by far the highest microplastics counts

of our processed samples. This trend could be explained both by the historical pollution present

in the Bladensburg watershed as well as ongoing construction occurring at the park with

insufficient barriers to block plastic debris from entering the watershed.

Policy Implications

The location of this research is particularly fixated at relevant areas for environmental

policy intiatives. From 1996 to present day, the Anacostia River has been listed by the Maryland

Department of the Environment as an impaired water for many reasons: trash, polychlorinated

biphenyls, total phosphorus, total nitrogen, biochemical oxygen demand, total suspended solids,

enterococcus, heptachlor epoxide, chloride, sulfate, habitat alterations, and a lack of riparian

buffer (Maryland’s Searchable Integrated Report Database [Combined 303(d)/305(b) List],

2022). Because of this, the Anacostia River requires a Total Maximum Daily Load (TMDL).

Since the Anacostia River watershed covers land in both Maryland and the District of Columbia,

a trash TMDL was created in cooperation between the Environmental Protection Agency Region

3, the District Department of the Environment, and Maryland Department of the Environment



67

(Total Maximum Daily Loads of Trash for the Anacostia River Watershed, Montgomery and

Prince George’s Counties, Maryland and the District of Columbia, 2010). The TMDL identifies

impaired waterways, sets targets to limit pollution, establishes guidelines for monitoring, and

summarizes trash reduction plans. This is an essential component of creating a collaborative,

interjurisdictional plan for preventing plastic entry into the Anacostia River watershed, but there

is an ongoing need to address the issue.

Creating a uniform methodology for quantifying or categorizing microplastics in

waterways is an essential foundation for making policy changes that mitigate plastic pollution.

The most impactful and cost-effective way of reducing environmental plastic pollution is through

source reduction and intervention before plastics enter the environment (Environmental

Protection Agency, 1990). This method is preferred over waste management and removal, as

plastic breaks down once it enters the environment and collecting and separating plastic from

other waste is a complicated process. Small countries with underdeveloped infrastructure for

waste management tend to burn plastics to reduce their volume, emitting a significant amount of

greenhouse gas and other poisonous gas in the process (Kibria et al., 2023). Therefore, plastics

not only pose their own challenges to environmental and human health, but their disposal can

also worsen the climate crisis.

Since secondary microplastics originate from weathered macroplastics, identifying where

microplastics exist in high concentrations and predicting their source material can inform policy

decisions to regulate plastic production. Data accessibility, quality, and sharing are three of the

top four challenges faced by state governments when using data in the policymaking process

(Bergh et al., 2018). However, having specific data and evidence to apply to an issue can make it

easy to pass effective legislation regarding the issue (Bergh et al., 2018). In addition, anti-litter



68

campaigns and litter-management solutions can be effectively targeted to specific communities

based on data about the types of litter that accumulate as well as the area in which it

accumulates. Carpenter & Wolverton (2017) utilized a behavioral archaeology approach to

understand the littering patterns in four sites along a stream in Texas and provide solutions. For

instance, they posit that in an area with a great amount of plastic that is high access, a locally

targeted campaign with signage both encouraging environmental stewardship and highlighting

the consequences of littering in conjunction with media coverage and greater educational

outreach would be the best strategy for reducing litter (Carpenter and Wolverton, 2017). In

contrast, to reduce litter at a site with fewer direct sources of litter and more lightweight plastic,

the study suggests plastic bag bans would yield the greatest effect. The site-specificity of plastic

introduction into local environments requires specific strategies to yield the greatest effects.

Thus, tailoring campaigns to reduce plastic waste towards local communities utilizing data from

analysis of local sites would be an effective strategy for diminishing the amount of plastic found

in local waterways.

One example of policymakers using local data to inform legislation has already occurred

on the local College Park scale. In early 2023, the College Park City Council unanimously

passed the Better Bag Bill, banning single-use plastic bags for College Park businesses (Wilson,

2023). Our preliminary research was cited during the hearing for the bill in College Park City

Hall (Artero, 2023), and council member Maria Mackie was motivated to vote for the bill to keep

plastic pollution out of the Anacostia watershed (Wilson, 2023). Similar initiatives have taken

place in other major cities and states that have passed policies banning or restricting various

forms of single-use plastics (Herberz et al., 2020). Plastic bag bans are generally supported by

local residents, but their effectiveness is contingent upon enforcement, as a survey conducted in



69

Ohio found that consumers are not self-motivated to stop using plastic bags (Bartolotta & Hardy,

2021). With store policies and legislation, however, consumers used reusable bags and still

supported businesses that enforced plastic bag taxes or did not offer plastic bags (Bartolotta &

Hardy, 2021). In the long-term, disposable plastic bag taxes consistently incentivize consumers

to use reusable bags and disincentivize plastic bag use, according to a study conducted in

Montgomery County, Maryland, where a plastic bag tax was passed in 2011 (Simon, 2020).

On the national level, the Organization for Economic Co-operation and Development is

directing policy efforts toward the regulation of microfibers shed from garments and tire and

road wear particles (Organization for Economic Co-operation and Development [OECD], 2021).

Emissions of these particles occur during several stages of their life cycles, allowing regulation

at various stages of production and use. Although preventative measures early in the production

processes are likely to be the most cost-effective of the solutions, improving waste management

for these products and containing them before they reach the environment are also areas of

possible regulation to diminish their presence in the environment. Regulating these particles is

made more complicated by a lack of consensus in the community that studies microplastics, as

well as a wide variation in sampling and characterization methods. This study seeks to rectify the

gap in methods by proposing a streamlined method for collecting, processing, and analyzing

freshwater microplastics samples. Standardizing the methods for collecting and processing

microplastics will allow easier comparison of results across areas, which will increase the impact

of data on policy making (OECD, 2021).

Much of the literature on plastics policy is international and focuses on marine

macroplastics, specifically revolving around those related to the Pacific Garbage Patch.

Policymakers are often prompted to pass legislation by seeing graphic photos of megafauna that



70

are negatively impacted by macroplastics (GESAMP, 2016). As a result, the lower visibility

plastics and impacts of plastic are not targeted as often because they do not provide the same

visual motivation. Regulating the production of larger plastics is an effective mechanism for

reducing the amount of microplastics in the environment, as many of these microplastics are

secondary microplastics. India targeted macroplastics through a single-use plastic ban and

adopted extended producer responsibility to manage waste from plastic production (Singh &

Biswas, 2023). To avoid economic losses and malpractice, lawmakers focused on investment in

finding compostable plastic alternatives to plastic items. Enforcement teams and border

checkpoints were also created to monitor the illegal manufacture, transport, and use of banned

single-use plastics (Singh & Biswas, 2023). A holistic literature review of the effectiveness of

plastic bag bans internationally found that plastic bag bans have been rendered ineffective by

lapses in monitoring and enforcement (Muposhi et al., 2022). They are similarly undermined by

corporations deflecting accountability to the consumer and national governments. The study also

found that a global treaty that standardizes the ban could aid in the enforcement of the bans and

prevent smuggling. Furthermore, another key to the effective bans is to encourage the voluntary

participation of consumers through education and the promotion of reusable shopping bags

(Muposhi et al., 2022).

The United Nations has various working groups dedicated to minimizing impacts from

anthropogenic plastic production and consumption, primarily Working Group 40 of the Joint

Group on the Scientific Aspects of Marine Environmental Protection (GESAMP). According to

this report, much of microplastic contribution can be traced back to consumer use; however,

upstream solutions enacted by producers are generally more cost-effective than downstream

solutions (GESAMP, 2016). In the same report, GESAMP highlighted the importance of



71

collaborating with national governments, which should focus on investing in infrastructure,

creating incentives for following government regulations, supporting research and development,

and encouraging enhanced producer and consumer responsibility. Governments should promote

private sector investment in green chemistry and other aspects of product design aimed at the

reduction of microplastic generation and waste. Local governments can focus on proper

wastewater treatment and port reception infrastructure to prevent microplastics from making

their way into the marine environment (GESAMP, 2016).

On the local scale in College Park, Bladensburg offers fishing and boating activities,

putting people in direct contact with the water that is downstream from construction sites.

Microplastic concentrations seem to peak during springtime and remain higher in the summer-

times when people are outside and more likely to come into contact with the water, especially

children. Thus, creating local campaigns to prevent littering would directly benefit highly

traversed sites, such as Bladensburg, where a great deal of litter comes from direct littering.

Government regulations to discourage the use of plastic and promote the development and use of

compostable products will diminish the damage caused by litter when these products inevitably

make their way into the environment.

Across freshwater microplastics, levels are highest in urban and estuarine streams (Luo et

al., 2019; Kye et al., 2023). A variety of land use covers our study sites, including designated

park areas, parking lots, and roadside areas buffered by trees. Of the three sites included, the

Iribe site was the closest to a major road, US Route 1 (see Figure 43). This factor is a way to

track the potential prevalence of tire-related microplastics in the region (Kole et al. 2017; Davey

et al., 2023). Collecting local data such as this will enable policymakers to create targeted



72

campaigns to reduce microplastics, such as by allowing them to determine if targeting consumers

or producers will yield the greatest change.

Figure 43: Distance from Major Roads between Bladensburg, Golf Course, and Iribe Sampling
Sites

Future Research Directions

This research contributes to the larger literature primarily in its focus on freshwater and

economical methods in microplastic detection. For further development, we suggest integration

of more computer vision assisted methods of quantifying microplastic concentration in samples,

and sampling in freshwater areas that are more impacted by tidal cycles. Additionally future

work should exclude the use of unfiltered potassium hydroxide pellets due to their plastics

contamination. We believe future sampling should occur at standard tidal influx times and

seasonality should be considered more robustly. We also recommend taking an optical density

measurement at each site to have a quantified measurement of turbidity rather than one based on

visual observations of the sample site, which could lead to better analysis of any relationship

between turbidity and microplastics. We believe that the use of a pre-staining procedure is useful

for limiting in-lab contamination from ambient sources. These ambient sources are important to



73

consider, especially when processing samples outside of a “clean room” setting as they have

been found to contribute particles with diameter < 50 µm (Zobkov et al., 2019; Kye et al., 2023).

We maintain an abundance of processed and filtered environmental samples which could be

counted using fluorescence microscopy. However, these filtered samples have the same

contamination issues as those counted in this report. As a result, future steps should involve

sample collection and processing, using our same methodology with the exception of digestion

with a pre-filtered concentrated potassium hydroxide solution.

A variety of questions remain in regard to microplastics which should be answered.

Related to our work, we believe that more extensive quantification is important in local streams

and waterways, especially in highly populated areas. We believe that our work helps to

troubleshoot microplastics methodology for these samples and lays the groundwork for more

effective quantification in the future.

Equity Impact Analysis

Establishing an equity impact statement for this research is crucial for ensuring that the

project has the intended social impacts and avoids undue harm to impacted communities. In

establishing this report, the first step is to identify stakeholders and currently or potentially

impacted communities. As members of an affluent campus community upstream from

communities threatened by environmental injustice, our project aims to quantify our

community’s contribution to microplastic quantities in the watershed in order to create a

foundation for policies addressing the issue.

Environmental Justice

Environmental pollutants such as microplastics pose a widespread threat once they enter

the environment. Preliminary research indicates that microplastics are prolific in the human



74

body. Microplastics have been identified in human blood, lung tissue, breastmilk, and placenta

(Goodman et al., 2021; Ragusa et al., 2022). Researchers have begun to report adverse effects of

microplastics on human health, such as respiratory inflammation and lesions, inhibition of cell

proliferation, oxidative stress, and an accumulation in cells (Blackburn & Green, 2022;

Goodman et al., 2021). While research defining the full extent of microplastic impacts on human

health is ongoing, the pervasiveness of these pollutants is enough to require expanded research

into microplastic quantities and concentrations in the environment. In order to figure out where

to target remediation efforts, we need to know how microplastics are distributed to focus our

efforts on the hardest hit areas. Additionally, we need to be able to compare results across studies

in order to understand the relative severity of the areas. By looking to create an accessible and

uniform collection and quantification procedure, we hope to enable cross comparison between

studies and help policymakers learn where to target remediation policies.

Our watershed of study covers communities with a variety of resources to support health

access, environmental safety, and climate change resilience. The two southernmost sampling

sites, NE Gauge and Bladensburg, are located in areas that face higher levels of environmental

injustice than the rest of the sites. According to the EPA’s EJScreen, NE Gauge and Bladensburg

face the highest levels of diesel particulate matter, traffic proximity, and superfund proximity,

among other environmental justice indices (US EPA, 2014). Additionally, the EJScreen identifies

that the NE Gauge sampling site is within the 95 - 100 percentile for lack of health insurance,

and the Bladensburg site is within the 90 - 95 percentile for the same. These are not the only

disparities these communities face, and the EJScreen does not quantify microplastics, but they

contribute to the larger, important narrative that communities downstream from College Park are

subject to higher levels of environmental injustice.



75

Microplastics within our watershed of study, the Anacostia watershed, affect the

community downstream in Bladensburg, the wider Washington, DC area, and all residents who

rely on food harvested or grown in the Chesapeake Bay region. Bladensburg, MD is an

environmental justice hotspot within Prince George’s County, which is 67% Black, faces some of

the highest levels of environmental injustice in the state (U.S. Census Bureau, 2021; Dosu et al.,

2019; Ieronimo, 2021). Marine litter and microplastics can be much more highly concentrated in

vulnerable communities, especially those that depend on wild seafood, so through quantifying

microplastics in our local watershed, we are providing the baseline information required to create

policies that will prevent environmental pollutants from entering the community (UN News,

2021).

We place particular emphasis on the potential health impacts of Washington, DC

residents consuming water collected from the Potomac River leading into the Chesapeake Bay.

While socioeconomically advantaged populations may have the opportunity to drink water or eat

food from outside sources that are free from microplastics, less affluent populations have little

autonomy over their exposure through ingestion. This only adds to the existing discrepancy

between urban versus non-urban environments, given urban environments consistently have

much higher concentrations of microplastics in water bodies and sediments. As a result, this

disparity of health opportunities creates inequities that may have far-reaching effects on overall

health and well-being. From an economic perspective, the majority of seafood harvesters living

in the Delmarva region are reliant on Chesapeake Bay organisms. Given how microplastics are

known to move up food webs and bioaccumulate in higher-level predators, like larger fish and

crabs, the influx of microplastics can have negative financial implications for an already

economically struggling population of seafood harvesters (Blackburn & Green, 2022).



76

As previously mentioned, this is just the tip of the iceberg and microplastics can impact

all populations. However, with these populations in mind, we define the primary way that a

positive impact for equity on this issue could be made is to (1) create an informative map about

microplastic exposure that is made available to the public and (2) suggest policy to the Maryland

Department of the Environment based on the findings from microplastic research (Maryland

Department of the Environment, 2023). Completing these goals will provide impacted

communities with easily accessible information on microplastics so that they can better influence

policy decisions. Finally, the policy shift must be written with equity and anti-discrimination

principles in mind to protect low-income communities and other vulnerable populations from

environmental hazards. Often, the burden of environmental laws can fall on low-income

populations, because they don’t have the resources to completely transform their consumption in

order to abide by the new policies. Instead of creating taxes and fees for the consumer, solutions

can focus on the producers of plastics via quotas and pollution control.

Accessible Protocols

Our project also creates the potential for the