Partitioning of Hydrophobic Organic Contaminants and Microbial Communities on Microplastics

W&M ScholarWorks

Dissertations, Theses, and Masters Projects Theses, Dissertations, & Master Projects

2019

Kelley Ann Uhlig William & Mary - Virginia Institute of Marine Science, [email protected]

Follow this and additional works at: https://scholarworks.wm.edu/etd

Part of the Environmental Sciences Commons, and the Marine Biology Commons

Recommended Citation Uhlig, Kelley Ann, "Partitioning of Hydrophobic Organic Contaminants and Microbial Communities on Microplastics" (2019). Dissertations, Theses, and Masters Projects. Paper 1563898639. http://dx.doi.org/10.25773/v5-wpxh-kx69

This Thesis is brought to you for free and open access by the Theses, Dissertations, & Master Projects at W&M ScholarWorks. It has been accepted for inclusion in Dissertations, Theses, and Masters Projects by an authorized administrator of W&M ScholarWorks. For more information, please contact [email protected].

Partitioning of Hydrophobic Organic Contaminants and Microbial Communities on Microplastics

A Thesis

Presented to

The Faculty of the School of Marine Science

The College of William and Mary in Virginia

In Partial Fulfillment

of the Requirements for the Degree of

Master of Science

Kelley Ann Uhlig

May 2019

APPROVAL PAGE

This thesis is submitted in partial fulfillment of

the requirements for the degree of

Master of Science

Kelley Ann Uhlig

Approved by the Committee, April 2019

Robert C. Hale, Ph.D. Committee Chair / Advisor

Kirk J. Havens, Ph.D.

Drew R. Luellen, Ph.D.

Bongkeun Song, Ph.D.

For my past and future self.

TABLE OF CONTENTS

DEDICATION ...... iv

ACKNOWLEDGEMENTS ...... vii

ABSTRACT ...... viii

GENERAL INTRODUCTION ...... 2

Plastics in the Marine Environment ...... 2

Fate & Effects ...... 5

Rationale ...... 9

Tables ...... 11

Figures ...... 12

References ...... 13

CHAPTER 1 – Comparative Partitioning Of Hydrophobic Organic Contaminants

To Petroleum- And Bio-Based Microplastics ...... 17

Introduction ...... 18

Materials and Methods ...... 21

Results and Discussion ...... 25

Conclusions ...... 31

Tables ...... 35

Figures ...... 43

References ...... 49

CHAPTER 2 – Characterizing the Microbial Diversity of Estuarine Biofilms on

Microplastics ...... 57

Introduction ...... 58

Materials and Methods ...... 62

Results ...... 66

Discussion ...... 71

Conclusions ...... 80

Tables ...... 82

Figures ...... 102

References ...... 107

POLICY CONSIDERATIONS ...... 114

Introduction ...... 114

Considerations for bio-based plastics ...... 115

Future considerations ...... 117

References ...... 119

VITA ...... 120

ACKNOWLEDGEMENTS

I would like to thank my advisor, Rob Hale, for taking a chance on me and allowing me the space to pursue my interests. Rob, you have been a constant pillar of support and understanding, an endless source of knowledge, and I couldn’t have done this without you. You’ve encouraged me to think critically, Second, I would like to acknowledge my committee members for their endless patience and support; Bongkeun Song, Drew Luellen, and Kirk Havens. BK, thank you for the guidance through the new and wonderful world of molecular biology, for taking the time to call me from Korea to help me sort through the mess I had made, your endless patience when my efforts did not seem to be paying off, and your advice and guidance for how to make an impact with this work. Thank you so much Drew for your knowledge and support with all things GC/MS, your no-nonsense attitude, and of course, for beer. None of this work would have happened without you. Last but not least, thank you Kirk for keeping things in perspective, for your encouragement and understanding, There are so many people at VIMS who have helped me over the years. At CCRM, thanks to Kory Angstadt & Dave Stanhope for providing my initial PHA stock, 3D printed slabs of plastic, and for always being great to talk with. Thank you to Jeff Sheild’s and Juan Pablo Huchin-Main for the blue crabs. To my south CBH 3rd floor family: Ellen Harvey, you are the rock of our lab! I wouldn’t have been able to find anything or even know where to start with this project without you and your keen advice. Mark LaGuardia, you’ve always been encouraging and supportive, I’m so glad you’re able to turn your years of hard work in a PhD! Matt Mainor, I always enjoyed our conversations about sailing, technology, you name it. George Vadas, thank you for the advice over the years and for loaning me several more separatory funnels so I that I could get my extractions done faster. Thank you to Mike Unger, for making me think more critically about science in genral. I continue to ask and will always ask, “What’s the question?” A huge thank you to Gail Scott for troubleshooting all of the things. My gratitude to the Song lab group: Samantha Fortin, Miguel Semedo, Ken Czapla, Stephanie Wilson, Renia Passie and Michele Cochran for answering so many of my questions and letting me invade your space. Thank you Vaskar Nepal for guiding me through PERMANOVA when I was so lost. Thank you to my partner in life and love, Chase Long, for always believing in me even when I didn’t believe in myself. Where would I be without you! I’m not sure but I would likely be hungry. I’m grateful to good friends, particularly my Tyndall Point family (Pamela Braff, Alison Bazylinski, and Alex Baruch), for making graduate school one of the best times in my life. Special thanks to Reviewer 1 from my EPA STAR application. Your harsh words telling me that my project was not worthy of a Master’s degree and that I would not succeed in an environmental science career were a constant reminder of why I had to accomplish the goals I laid out for myself.

vii

ABSTRACT

Microplastic contamination of aquatic environments has only recently caught the attention of scientists, regulators and the public. Microplastics are typically more recalcitrant than naturally occurring polymers and so have the potential to cause a range of issues, including increased exposure of marine life to chemical contaminants sorbed to or leached from microplastics, negative impacts due to ingestion of microplastics by biota, and the potential to carry and transport pathogenic and invasive species long distances. Bio-based, bio-degradable polymers have begun to gain market share as an alternative to traditional petrochemical-based plastics, but not much is known about their impacts in marine environments. The overall objective of this thesis was to improve our understanding of how bio-based microplastics compare to petrochemical-based plastics in the marine environment. This information could be used to evaluate the overall sustainability of bio-based polymers as replacements for petrochemical-based polymers. The first chapter of this work investigated the potential of four types of microplastics, polyethylene (PE), polyvinyl chloride (PVC) and two bio-based polymers, poly-3-hydroxybutyrate (PHB) and polylactic acid (PLA), to sorb hydrophobic organic contaminants (pyrene, PCB-153, and BDE-47) from the surrounding water column. It also examined how co-exposure to several of these contaminants influenced their sorption. The bio-based polymers used in this work, exhibited lower affinity for the organic contaminants investigated compared to the more widely used, petrochemical- derived microplastics. This may be due to several factors including hydrophobicity of the plastic surfaces and the chemical structure of each plastic. Further, competition between several co-exposed contaminants led to an overall decrease in chemical partitioning on polyethylene microplastics. The second chapter reported on the microbial composition of biofilm communities that form on bio-based (PHB and PLA) and petrochemical-based (PE and PVC) microplastics in comparison to a naturally occurring polymer, chitin. Microbial compositions of biofilms that formed on the different plastics were similar during the first and second week of growth, but chitin exhibited a distinct community from the microplastics. By the fourth week of growth, all substrates had a similar community composition. Diversity was generally higher on bio-based plastics. Genera harboring marine pathogens and hydrocarbon-degrading bacteria were identified on all substrates. This work has implications to policy surrounding marine debris issues, exploring the more nuanced differences between bio-based polymers and petrochemical polymers, introducing concerns over additive use in bio-based polymers, and reinforcing the need for “eco-cyclable” materials in single-use items.

viii Sorption of Hydrophobic Organic Contaminants and Microbial Communities on Microplastics

GENERAL INTRODUCTION

1. Plastics in the Marine Environment

The subject of plastic pollution in the marine environment is not new. The first documented account of plastic debris at sea was reported in 1972 by Carpenter and

Smith. Plastics were collected during 11 tows of a 1 m, 330 µm mesh-size neuston net through the Sargasso Sea. They provided estimates regarding the density and the weight of plastics, reporting a mean of 3537 particles of plastic per km2, corresponding to 286.8 g per km2. Around the same time, Scott (1972) reported an abundance of plastic pollution littering coastlines. Rothstein (1973) found evidence of seabirds ingesting plastic pollution as early as 1964. These dates follow the start of the boom in plastics manufacturing and use in the1950s (Jambeck et al. 2015). Carpenter and Smith predicted that the “increasing production of plastics, combined with present waste disposal practices, will probably lead to greater concentrations on the sea surface.” That has indeed occurred.

Plastics have been found in every body of water examined to date, from polar seas to remote mountain lakes (Obbard et al. 2014). Hence it is reasonable to conclude that plastic pollution is ubiquitous. Plastics enter the oceans from land sources by rivers, storm runoff and wastewaters, and catastrophic events; and from coastal and open ocean sources like lost cargo, derelict fishing and aquaculture gear, and trash left behind by beachgoers. Once water-borne, plastics are at the mercy of physical forcing. Floating plastics can strand and accumulate on shorelines due to wave, tidal, and wind action.

Plastics that are more dense than seawater concentrate closer to sources, such as in coastal and estuarine sediments, although recently these have been detected in deep ocean

sediments (Van Cauwenberghe et al. 2013). In the open ocean, elevated concentrations of floating plastic are found in the five oceanic gyres, where wind-driven Ekman convergence focuses debris (Law et al. 2014). Accumulation in the gyres can happen extremely rapidly, as demonstrated in a study by Law and colleagues using drifter data from 1989 to 2009. The authors found that a drifter originating from Washington, D.C. might reach the North Atlantic subtropical gyre in as little as 40 days (Law et al. 2010).

Furthermore, microplastics can be distributed throughout the mixed layer by vertical mixing (Kukulka et al. 2012) and altered buoyancy from biofouling (Ye & Andrady

1991; Lobelle & Cunliffe 2011). Evidence also suggests that sediments are a ‘sink’ for plastic debris (Goldberg 1997), including deep-sea sediments of the Southern Ocean and the Porcupine Abyssal Plain (Van Cauwenberghe et al. 2013).

Defining ‘plastic’ presents its own complexity. The Oxford Dictionary (2019) defines plastic as “a synthetic material made from a wide range of organic polymers…that can be molded into shape while soft, and then set into a rigid or slightly elastic form.” Organic polymers are chemically linked repeating units of hydrogen and carbon, though other elements may be present as well. It is useful to remember that all plastics are polymers, but not all polymers are plastics (e.g. deoxyribonucleic acid (DNA)). The term plastic also refers to the property of plasticity, the ability to deform without breaking. Leo

Baekeland coined the term ‘plastics’ in 1907 when he invented the first synthetic plastic, a phenolic resin branded Bakelite (Helmenstine 2018).

1.1 Petroleum-Based Plastics

Most plastics are produced from petroleum-based polymers. The naptha fraction from crude oil distillation serves as the feedstock for polymerization. Monomers of ethylene

and propylene react with catalysts to form the polymeric chains constituting plastics

(PlasticsEurope 2019). Polyethylene (PE), polypropylene (PP), polystyrene (PS), and polyvinyl chloride (PVC) are the plastics most commonly manufactured and hence, the most commonly found in aquatic environments. In 2014, global production of plastics topped 300 million metric tons (Mt). China, Europe, and the countries of the North

Atlantic Free Trade Agreement (Canada, United States, Mexico) produced over 60% of that plastic. Table 1 lists European plastics demand by polymer, which is representative of the rest of the developed world (PlasticsEurope 2018).

1.2 Bio-Based Plastics

Bio-based plastics are comprised of polymers derived from biomass (i.e. plants, animals, microbes), or non-fossilized organic material. Kabasci (2013) notes that the term

‘bioplastics’ is often not used in reference to the origin of the polymer, but rather to some functionality of the plastic; usually biodegradability (i.e. capable of being decomposed by bacteria or other living organisms). ‘Biopolymers’ is another term that is often confused with bio-based polymers. It refers to polymers synthesized by living organisms, such as polysaccharides, proteins, or nucleic acids. Some biopolymers can be extracted and processed for use in commercial products as plastics (i.e. polyhydroxyalkanoates;

Kabasci 2013). Figure 1 provides an overview of bio-based plastics.

Though bio-based plastics currently make up only 2% of the total polymer market, they have a presence in every sector, from rigid and flexible packaging to textiles and consumer goods (Chinthapalli et al. 2019). Worldwide production of bio-based plastics is predicted to keep pace with market share, growing at 4% annually from 7.5 Mt in 2018 to nearly 10 Mt by 2023 (Chinthapalli et al. 2019). The two bio-based polymers I focus on

for this work are polyhydroxyalkanoate (PHA) and polylactic acid (PLA). PHA and PLA are the main drivers of this growth in bio-based, biodegradable plastics. The production capacities for PHA and PLA are expected to grow nearly 500% and 50% by 2023, respectively (European Bioplastics 2018).

2. Fate & Effects

Plastic products are typically engineered to be environmentally stable. This derives from their long chain olefin-derived structures, as well as the introduction of specific chemical additives, such as UV blockers and microbicides. Degradation half-lives may range from months to centuries, depending on composition and environmental conditions.

Modes of degradation may be chemical: e.g. UV induced photo-oxidation, thermo- degradation, hydrolysis, and biodegradation. Or it may be physical, such as fragmentation due to abrasion. Degradation of polymers is characterized by a decrease in molecular weight, a loss of the mechanical properties (i.e. tensile strength, compression, and impact strength), or an alteration of surface characteristics (i.e. discoloration, cracking and

“chalking”). Alterations may be monitored by spectroscopic techniques such as Fourier- transform infrared (FTIR) or Raman spectroscopy (Andrady 2015).

Biodegradation of most plastics in the marine environment is slow, though formation of a biofilm may be rapid (Lobelle & Cunliffe 2011; Ho et al. 1999). Many organisms that form biofilms also secrete enzymes that can degrade petrochemical-based plastics.

However, the slow kinetics of this process limit elimination of plastic debris from the oceans (Andrady 2015).

2.1 Chemical Interactions

A major concern regarding plastic pollution is the potential for deleterious consequences on organisms that ingest the particles. These may arise from chemicals that concentrate on their surfaces from the surrounding water (sorption) and from additives that may leach from within the plastic itself. The present study is concerned exclusively with sorption of contaminants. The partitioning of organic chemicals to a solid, is a complex process that is dependent on the physico-chemical properties of the solid and the chemical of interest. Sorption is a physical and chemical behavior between a surface and a chemical. Specific surface area of the solid and Van der Waals’ forces (i.e. London dispersion and dipole interactions) acting between its surface and the chemical dominate the physical aspect of sorption. Sorption to a polymer is dictated by its lipophilicity and the aqueous solubility, or octanol-water partitioning coefficient of the chemical. (Delle

Site 2001; Wang et al. 2016)

Accordingly, polymers such as PE have found utility as single-phase passive samplers of hydrophobic organic contaminants (HOCs) in aqueous systems. This is due to their low cost, ease of analysis, and suitability for monitoring a variety of HOCs (Perron et al.

2013). Their diffusivity and crystallinity also make some polymers excellent sorbers

(Wang et al. 2016). Many polymers are composed of crystalline and amorphous regions, characterized by their level of molecular organization. The crystalline region is comprised of a well-organized lattice structure of monomer units, and the amorphous region exhibits a more random arrangement of molecules (Wang et al. 2016). Teuten et al. (2009) hypothesized that increased sorption of HOCs occurs in the amorphous region of the polymer structure, where the polymer exhibits more viscous liquid properties

amenable to sorption. The glass transition temperature (T") is the temperature below which a given polymer is considered more “glassy,” or crystalline (Xing and Pignatello

1997; Teuten et al. 2009). This glass transition temperature can inform us about the likely propensity of a polymer to sorb of HOCs. Additionally, hydrophobicity of a surface can be measured as the contact angle between a drop of water and the polymer surface (Yuan

& Lee 2013). The hydrophobic effect is the observed tendency for nonpolar substances in solution to aggregate to exclude water, maximizing hydrogen bonds between water molecules and minimizing interactions with nonpolar molecules. The greater the hydrophobicity, the greater the effect. Therefore, a more hydrophobic surface should lead to greater sorption of HOCs. Furthermore, each HOC varies in its hydrophobic nature, measured by its solubility in water (#) and octanol-water partitioning coefficient (K%&).

The direct correlation between HOC accumulation in n-octanol and accumulation in the fatty tissues of organisms, makes K%& an excellent proxy to estimate bioconcentration.

That is, the higher the K%& of the HOC, the greater the likelihood of accumulation of that

HOC within an animal.

2.2 Competitive Sorption

Hydrophobic organic contaminants are present as complex mixtures in the natural environment. Competition between these contaminants for sorption can alter the partitioning of HOCs to polymers. Competitive sorption between organic pollutants has been well documented in soils and sediments (Xing et al. 1996; Zhao et al. 2002, Yu et al. 2005), with some evidence for increased bioavailability of the competing sorbents

(White et al. 1999). Xing and Pignatello first investigated competitive sorption of dichlorobenzene to PVC microspheres to help explain sorption to soil organic matter

(1997). They found that the addition of chlorobenzene as a co- solute suppressed the distribution coefficient of dichlorobenzene to PVC, and suggested that competition between structurally similar molecules to glassy polymers may be enhanced due to the co-solute blocking “holes” in the crystal structure where dichlorobenzene would normally sorb. There has also been evidence of competitive sorption between HOCs vying for the same surface on ultra-high molecular weight (UHMW) PE and PVC microplastics.

Specifically, Bakir et al. (2012) observed that DDT suppressed the sorption of phenanthrene.

2.3 Microbial Diversity

A diverse array of microbial life colonizes marine submerged surfaces, and plastic debris is no exception (Carson et al. 2013; Zettler et al. 2013; McCormick et al. 2014;

Oberbeckmann et al. 2016). Colonization of a surface provides bacteria with several important advantages, including greater access to nutrients, physical stability and shelter from predation (Dang & Lovell 2000). Colonization by microorganisms also affects the fate of plastic debris by modifying the density of floating plastics (Ye & Andrady 1991), initiating biodegradation (Sudhakar et al. 2007; Artham et al. 2009; Muthukumar et al.

2011; Zettler et al. 2013), altering chemical sorption and leaching (Dang & Lovell 2000;

Froehner et al. 2012), and influencing ingestion of debris by organisms (Carson et al.

2013). Furthermore, pathogenic, toxic, and invasive species have been found to colonize plastic debris (Zettler et al. 2013; Oberbeckmann et al. 2016). Marine-borne plastics are more persistent than most natural polymeric debris found in coastal and open-ocean systems, potentially allowing for greater transport and distribution of associated colonizing microbes. Hence, increasing levels of such plastics in the environment may

enhance the growth and distribution of opportunistic organisms and threaten ecological systems. Ingestion of such plastics may also impact normal resident gut microbiomes that are critical for organismal health. Not all plastics are created equal though, as the ingestion and subsequent breakdown of PHAs to short-chain fatty acids has been shown to impart a level of immunity to pathogenic bacteria (Liu et al. 2010; Sui et al. 2012).

Phylogenetic analysis utilizing high-throughput sequence analysis of 16S ribosomal

RNA (rRNA) genes has been used to describe microbial communities on plastics collected from the open ocean (Zettler et al. 2013) and an urban river environment

(McCormick et al. 2014). Others have used these techniques to investigate changes in microbial community over time on plastics buried in coastal sediment microcosms

(Harrison et al. 2014), and PET water bottles deployed in the North Sea (Oberbeckmann et al. 2016). These techniques involve the extraction of DNA from the plastics, followed by polymerase chain reaction (PCR) amplification of rRNA genes and creation of a clone library. Once cloned, these libraries can be screened using probes to detect specific organisms or groups of organisms. Sequences can be further filtered for comparison using software such as mothur (Schloss et al. 2009) or the R packages DADA2 (Callahan et al. 2016) and phyloseq (McMurdie & Holmes 2013).

3. Rationale

This study addresses gaps in knowledge related to the sorption of HOCs to bio-based plastics. In a review of the literature, I was only able to identify one study focusing on this. Matsuzawa et al. (2010) investigated the association of dissolved HOCs with several biodegradable polymers, but the study was not intensive and did not calculate sorption isotherms or partitioning coefficients. They did, however, underline the potential

importance of sorption to bio-based plastics. My research sheds light on the sorption of

HOCs to two bio-based plastics with substantial market shares. These shares may increase dramatically in the future as consumers and policy influence manufacturers.

Competitive sorption has only been investigated for microplastics explicitly in one study by Bakir et al. (2012). They utilized only two types of polymers (UHMW PE and

PVC) and two HOCs. Despite being quoted extensively as a need and consideration for research purposes, further studies have yet to be published. My work further investigates the question of competitive sorption with different polymers and different sorbents.

Though individual microbes implicit in degradation have been identified for several bio-based plastics and traditional petrochemical plastics, there is a lack of understanding if these microbes are present in natural environments. Harrison et al. (2011) identified three research needs concerning microbial interactions with marine plastics, and my

Master’s thesis work addresses the second: “Wider research into the colonization and biodegradation of synthetic polymers by microorganisms.”

4. Tables

Table 1. European plastics demand by polymer type

Plastic Demand Example

food packaging, hinged caps, bank notes, pipes, PP 19.3% automotive parts insulated coatings (PTFE), hub caps (ABS), roofing Others 19.0% (PC), optical fibers (PBT), touch screens (PMMA) food packaging films, film bags, reusable bags, trays and LDPE, LLDPE 17.5% containers HDPE, MDPE 12.3% toys, milk bottles, pipes, shampoo bottles, housewares PVC 10.2% window frames, flooring, pipes PUR 7.7% mattresses, insulation panels, foam cushioning eye-glass frames, plastic cups, packaging, building PS, EPS 6.6% insulation PET 6.4% bottles for water, soft drinks, juice, cleaners, etc.

PP, polypropylene; PTFE, polytetrafluoroethylene; ABS, acrylonitrile butadiene styrene; PC, polycarbonate; PBT, polybutylene terephthalate; PMMA, polymethyl methacrylate; LDPE, low density polyethylene; LLDPE, linear low density polyethylene; HDPE, high density polyethylene; MDPE, medium density polyethylene; PVC, polyvinylchloride; PUR, polyurethane; PS, polystyrene; EPS, expandable polystyrene; PET, polyethylene terephthalate. Modified from PlasticsEurope 2018.

5. Figures

Bio-based plastics

Based on natural polymers Polymerized from Proteins bio-based monomers Lignin

Natural rubber Polysaccharides Monomers from Monomers from chemical synthesis biochemical synthesis

Isosorbide, Succinic acid, furanics, etc. Structural Storage lactic acid, etc. Others Cellulose, Starch chitin, etc. PHAs

Figure 1. Overview of bio-based plastics outlining the different types, with examples.

Modified from Kabasci (2013).

6. References

Andrady, A. L. (2015). Persistence of Plastic Litter in the Oceans. In M. Bergmann, L. Gutow, & M. Klages (Eds.), Marine Anthropogenic Litter (pp. 57–72). https://doi.org/10.1007/978-3-319-16510-3_3

Artham, T., Sudhakar, M., Venkatesan, R., Madhavan Nair, C., Murty, K. V. G. K., & Doble, M. (2009). Biofouling and stability of synthetic polymers in sea water. International Biodeterioration & Biodegradation, 63(7), 884–890. https://doi.org/10.1016/j.ibiod.2009.03.003

Bakir, A., Rowland, S. J., & Thompson, R. C. (2012). Competitive sorption of persistent organic pollutants onto microplastics in the marine environment. Marine Pollution Bulletin, 64(12), 2782–2789. https://doi.org/10.1016/j.marpolbul.2012.09.010

European Bioplastics (2018). Bioplastics market data 2018. Retrieved from www.european-bioplastics.org

Callahan, B. J., McMurdie, P. J., Rosen, M. J., Han, A. W., Johnson, A. J. A., & Holmes, S. P. (2016). DADA2: High-resolution sample inference from Illumina amplicon data. Nature Methods, 13(7), 581–583. https://doi.org/10.1038/nmeth.3869

Carpenter, E. J., & Smith Jr, K. L. (1972). Plastics on the Sargasso Sea surface. Science, 175(4027), 1240. Retrieved from http://www.jstor.org/stable/1733709

Carson, H. S., Nerheim, M. S., Carroll, K. a., & Eriksen, M. (2013). The plastic- associated microorganisms of the North Pacific Gyre. Marine Pollution Bulletin, 75(1–2), 126–132. https://doi.org/10.1016/j.marpolbul.2013.07.054

Chinthapalli, R., Skoczinski, P., Carus, M., Baltus, W., Guzman, D. de, Käb, H., … Ravenstijn, J. (2019). Bio-based Building Blocks and Polymers Global Capacities and Trends 2016–2021.

Dang, H., & Lovell, C. R. (2000). Bacterial primary colonization and early succession on surfaces in marine waters as determined by amplified rRNA gene restriction analysis and sequence analysis of 16S rRNA genes. Applied and Environmental Microbiology, 66(2), 467–475. https://doi.org/10.1128/AEM.66.2.467-475.2000

Delle Site, A. (2001). Factors affecting sorption of organic compounds in natural sorbent/water systems and sorption coefficients for selected pollutants: A review. Journal of Physical and Chemical Reference Data, 30(1), 187–439. https://doi.org/10.1063/1.1347984

Froehner, S., Machado, K. S., Dombroski, L. F., Nunes, A. C., Kishi, R. T., Bleninger, T., & Sanez, J. (2012). Natural biofilms in freshwater ecosystem: Indicators of the

presence of polycyclic aromatic hydrocarbons. Water, Air, & Soil Pollution, 223(7), 3965–3973. https://doi.org/10.1007/s11270-012-1164-y

Goldberg, E. D. (1997). Plasticizing the seafloor: An overview. Environmental Technology, 18(2), 195–201. https://doi.org/10.1080/09593331808616527

Harrison, J. P., Sapp, M., Schratzberger, M., & Osborn, A. M. (2011). Interactions between microorganisms and marine microplastics: a call for research. Mar. Technol. Soc. J., 45(2), 12.

Ho, K.-L. G., Pometto III, A. L., & Hinz, P. N. (1999). Effects of temperature and relative humidity on polylactic acid plastic degradation. Journal of Polymers and the Environment, 7(2), 83–92. https://doi.org/10.1023/A:1021808317416

Jambeck, J. R., Geyer, R., Wilcox, C., Siegler, T. R., Perryman, M., Andrady, A., … Law, K. L. (2015). Plastic waste inputs from land into the ocean. Science, 347(6223), 768–771. https://doi.org/10.1126/science.1260352

Kabasci, S. (Ed.). (2013). Bio-Based Plastics. https://doi.org/10.1002/9781118676646

Law, K. L., Morét-Ferguson, S. E., Goodwin, D. S., Zettler, E. R., DeForce, E., Kukulka, T., & Proskurowski, G. (2014). Distribution of surface plastic debris in the eastern Pacific Ocean from an 11-year data set. Environmental Science & Technology, 48(9), 4732–4738. https://doi.org/10.1021/es4053076

Law, K. L., Moret-Ferguson, S., Maximenko, N. A., Proskurowski, G., Peacock, E. E., Hafner, J., & Reddy, C. M. (2010). Plastic accumulation in the North Atlantic subtropical gyre. Science, 329(5996), 1185–1188. https://doi.org/10.1126/science.1192321

Liu, Y., De Schryver, P., Van Delsen, B., Maignien, L., Boon, N., Sorgeloos, P., … Defoirdt, T. (2010). PHB-degrading bacteria isolated from the gastrointestinal tract of aquatic animals as protective actors against luminescent vibriosis. FEMS Microbiology Ecology, 74(1), 196–204. https://doi.org/10.1111/j.1574- 6941.2010.00926.x

Lobelle, D., & Cunliffe, M. (2011). Early microbial biofilm formation on marine plastic debris. Mar. Pollut. Bull., 62(1), 197–200. https://doi.org/10.1016/j.marpolbul.2010.10.013

Matsuzawa, Y., Kimura, Z.-I., Nishimura, Y., Shibayama, M., & Hiraishi, A. (2010). Removal of hydrophobic organic contaminants from aqueous solutions by sorption onto biodegradable polyesters. Journal of Water Resource and Protection, 02(03), 214–221. https://doi.org/10.4236/jwarp.2010.23024

McCormick, A., Hoellein, T. J., Mason, S. A., Schluep, J., & Kelly, J. J. (2014). Microplastic is an abundant and distinct microbial habitat in an urban river. Environmental Science & Technology, 48(20), 11863–11871. https://doi.org/10.1021/es503610r

McMurdie, P. J., & Holmes, S. (2013). phyloseq: An R package for reproducible interactive analysis and graphics of microbiome census data. PLoS ONE, 8(4), e61217. https://doi.org/10.1371/journal.pone.0061217

Muthukumar, T., Aravinthan, A., Lakshmi, K., Venkatesan, R., Vedaprakash, L., & Doble, M. (2011). Fouling and stability of polymers and composites in marine environment. International Biodeterioration & Biodegradation, 65(2), 276–284. https://doi.org/10.1016/j.ibiod.2010.11.012

Obbard, R. W., Sadri, S., Wong, Y. Q., Khitun, A. a., Baker, I., & Thompson, R. C. (2014). Global warming releases microplastic legacy frozen in Arctic Sea ice. Earth’s Future, 2(6), 315–320. https://doi.org/10.1002/2014EF000240

Oberbeckmann, S., Osborn, A. M., & Duhaime, M. B. (2016). Microbes on a bottle: Substrate, season and geography influence community composition of microbes colonizing marine plastic debris. PLOS ONE, 11(8), e0159289. https://doi.org/10.1371/journal.pone.0159289

Perron, M. M., Burgess, R. M., Suuberg, E. M., Cantwell, M. G., & Pennell, K. G. (2013). Performance of passive samplers for monitoring estuarine water column concentrations: 1. Contaminants of concern. Environmental Toxicology and Chemistry / SETAC, 32(10), 2182–2189. https://doi.org/10.1002/etc.2321

PlasticsEurope. (2018). Plastics - The Facts 2018. PlasticsEurope. https://doi.org/10.1016/j.marpolbul.2013.01.015

PlasticsEurope. (2019). How are plastics made? Retrieved April 20, 2019, from https://www.plasticseurope.org/en/about-plastics/what-are-plastics/how-plastics-are- made

Rothstein, S. I. (1973). Plastic particle pollution of the surface of the Atlantic Ocean: Evidence from a seabird. The Condor, 75(3), 344–345. https://doi.org/10.2307/1366176

Schloss, P. D., Westcott, S. L., Ryabin, T., Hall, J. R., Hartmann, M., Hollister, E. B., … Weber, C. F. (2009). Introducing mothur: Open-source, platform-independent, community-supported software for describing and comparing microbial communities. Applied and Environmental Microbiology, 75(23), 7537–7541. https://doi.org/10.1128/AEM.01541-09

Scott, P. G. (1972). Plastics packaging and coastal pollution. International Journal of Environmental Studies, 3(1–4), 35–36. https://doi.org/10.1080/00207237208709489

Sudhakar, M., Trishul, A., Doble, M., Suresh Kumar, K., Syed Jahan, S., Inbakandan, D., … Venkatesan, R. (2007). Biofouling and biodegradation of polyolefins in ocean waters. Polymer Degradation and Stability, 92(9), 1743–1752. https://doi.org/10.1016/j.polymdegradstab.2007.03.029

Sui, L., Cai, J., Sun, H., Wille, M., & Bossier, P. (2012). Effect of poly-β- hydroxybutyrate on Chinese mitten crab, Eriocheir sinensis, larvae challenged with pathogenic Vibrio anguillarum. Journal of Fish Diseases, 35(5), 359–364. https://doi.org/10.1111/j.1365-2761.2012.01351.x

Teuten, E. L., Saquing, J. M., Knappe, D. R. U., Barlaz, M. A., Jonsson, S., Bjorn, A., … Takada, H. (2009). Transport and release of chemicals from plastics to the environment and to wildlife. Philosophical Transactions of the Royal Society B: Biological Sciences, 364(1526), 2027–2045. https://doi.org/10.1098/rstb.2008.0284

Van Cauwenberghe, L., Vanreusel, A., Mees, J., & Janssen, C. R. (2013). Microplastic pollution in deep-sea sediments. Environmental Pollution, 182, 495–499. https://doi.org/10.1016/j.envpol.2013.08.013

Wang, J., Tan, Z., Peng, J., Qiu, Q., & Li, M. (2016). The behaviors of microplastics in the marine environment. Marine Environmental Research, 113, 7–17. https://doi.org/10.1016/j.marenvres.2015.10.014

Xing, B., & Pignatello, J. J. (1997). Dual-mode sorption of low-polarity compounds in glassy poly(vinyl chloride) and soil organic matter. Environmental Science & Technology, 31(3), 792–799. https://doi.org/10.1021/es960481f

Xing, B., Pignatello, J. J., & Gigliotti, B. (1996). Competitive sorption between atrazine and other organic compounds in soils and model sorbents. Environmental Science & Technology, 30(8), 2432–2440. https://doi.org/10.1021/es950350z

Ye, S., & Andrady, A. L. (1991). Fouling of floating plastic debris under Biscayne Bay exposure conditions. Marine Pollution Bulletin, 22(12), 608–613. https://doi.org/10.1016/0025-326X(91)90249-R

Yu, Z., & Huang, W. (2005). Competitive sorption between 17α-ethinyl estradiol and naphthalene/phenanthrene by sediments. Environmental Science & Technology, 39(13), 4878–4885. https://doi.org/10.1021/es048558k

Zettler, E. R., Mincer, T. J., & Amaral-Zettler, L. A. (2013). Life in the “Plastisphere”: Microbial communities on plastic marine debris. Environmental Science & Technology, 47(13), 130619162220002. https://doi.org/10.1021/es401

Comparative Partitioning Of Hydrophobic Organic Contaminants To Petroleum- And Bio-Based Polymers

1. Introduction

Increasing reliance on plastic products since 1950 coupled with inadequate waste management practices has led to the escalating accumulation of plastic debris throughout the environment (Jambeck et al. 2015). Polymers produced in the greatest amounts are also the most common in waterways, i.e. polyethylene (PE), polypropylene (PP), polystyrene (PS), and polyvinyl chloride (PVC). Over time plastics weather and fragment and those < 0.5 mm in size are deemed “microplastics.” Many plastics persist for decades and are subsequently distributed in surface waters, stranded on coastlines, ingested by organisms and incorporated into sediments.

One path to reduce impacts is the substitution of sustainable, biodegradable plastics for petrochemical-based non-degradable plastics. Production of bio-based polymers, whose building blocks are sourced from non-fossilized biomass, has nearly doubled from

2.8 million in 2011 to 4.6 million tonnes in 2017, and is expected to top 5.0 million tonnes by 2022 (Aeschelmann & Carus 2015). Bio-based polymers are also an attractive alternative to petrochemical-derived plastics because renewable feedstocks are not at the mercy of volatile oil prices. Their production results in reduced greenhouse gas emissions and more diverse end-of-life options are available to process discarded bio-polymer products (InnProBio 2016). Though many bio-based polymers are biodegradable, this is not universally true. For instance, plant-based polyethylene terephthalate (bio-PET) bottles are no more biodegradable than conventional PET.

Degradability of conventional and bio-based polymers vary, and most are not easily assimilated into the carbon cycle. McDevitt et al. (2017) recently suggested a new standard of “Ecocyclable,” to designate materials that “degrade into products that are readily

incorporated into the natural carbon cycle, are nontoxic, and do not lead to the accumulation of persistent chemicals.” One such material is poly-3-hydroxybutyrate

(PHB), a bio-based polymer produced in situ by bacteria. This polymer has been successfully implemented in degradable crab trap escape panel, reducing ‘ghost fishing’

(Havens et al. 2015). Another potential option is polylactic acid (PLA), which is produced from the polymerization of lactic acid from fermented biomass (Groot et al. 2010). PLA is already widely used as an agricultural covering. Degradability is a factor of the environment in which a plastic resides, with slower rates in the open ocean (lower temperatures, reduced UV exposure) and faster rates on shore (physical breaking, elevated temperatures, high UV exposure) (Andrady 2015; McDevitt et al. 2017).

In addition to environmental persistence, another concern surrounding marine plastic debris stems from the potential of aquatic-borne plastics to concentrate hydrophobic organic contaminants (HOCs) from the surrounding environment (Mato et al. 2001; Teuten et al. 2007; Van et al. 2012; Rochman et al. 2013a; Rochman et al. 2013b). As plastic debris fragments, its surface area increases. This in concert with the greater hydrophobicity of some polymers, allows for higher concentrations of HOCs to sorb from the surrounding waters than to natural organic matter (Teuten et al. 2007). These fragments are also more readily ingested by the small organisms that make up the base of the food web (Teuten et al. 2007; Van et al. 2012; Rochman et al. 2013a; Rochman et al. 2013b). The debate remains unsettled concerning the potential for increased organismal exposure and bioaccumulation of HOCs (with resonating food web effects on top consumers – e.g. humans) from plastic debris. Some research suggests that plastics in the aquatic environment act as HOC sinks, decreasing bioavailability by reducing their free-water

column concentrations, and that HOCs do not readily desorb into organisms from ingested plastic (Gouin et al. 2011; Koelmans et al. 2016; Diepens & Koelmans 2018; Kleinteich et al. 2018).

The monomer units that create a polymer exhibit varied physical and chemical properties. Thus, it is unlikely that PE acts similarly to PHB in terms of capacity to sorb environmental contaminants. In any event, bio-based plastics are being promoted as an environmentally benign alternative to fossil-fuel-based plastics and are becoming a more common constituent of marine debris. Therefore, a better understanding of their interaction with HOCs in marine environments is needed. Only a few studies (Houghton & Quarmby

1999; Matsuzawa et al. 2010) have investigated the tendency for bio-based polymers to sorb water-borne contaminants. Hence further study to investigate how bio-based polymers may compare to conventional polymers is needed.

The phenomenon of competitive HOC sorption has been documented in soils

(McGinley et al. 1993; Xing et al. 1996), sediments (McGinley et al. 1993; Weber, Jr. et al. 2002, Zhao et al. 2002), carbon nanotubes (Yang et al. 2006; Wang et al. 2009; Shen et al. 2014), and other media (Sander & Pignatello 2005; Wang et al. 2006; Wang et al.

2007; Valderrama et al. 2010). In contrast, interactions with plastic debris have received only modest attention (Xing & Pignatello 1997; Li & Werth 2001; Bakir et al. 2012). It is crucial to assess how contaminant mixtures partition between polymers and the surrounding environment in order to understand how competitive sorption influences the fate and effects of HOCs. Multi-solute experiments can provide mechanistic details for sorption processes and insights into altered bioavailability of contaminants. There is some evidence that competition between solutes leads to increased bioavailability of displaced,

previously sorbed HOCs (White et al. 1999, Hummel et al. 2017). Competitive sorption could also have an impact on the use of plastic debris as passive samplers (Mato et al.

2001), as the presence of multiple HOCs could alter estimates of ambient contaminant levels.

The current study determines the partitioning coefficients of three HOC classes of concern (PAH, PCB, and PBDEs) to two traditional petrochemical-based polymers (high- density PE and PVC) and two bio-based polymers (PHB and PLA). We hypothesized that partitioning to the bio-based polymers would be significantly less than to more hydrophobic fossil-fuel derived polymers. Additionally, we investigated if the presence of multiple HOCs in solution altered their respective partitioning to PE by comparing partitioning coefficients of single and multi-solute HOC mixtures. We anticipated no change in the partitioning behavior of the contaminants in a complex mixture due to the perceived availability of sorptive sites on the polymer surface.

2. Materials and Methods

2.1 Materials

Solvents used were HPLC grade or better. Pyrene (Pyr), deuterated pyrene (d10-Pyr),

2,2’,4,4’,5,5’-hexachlorobiphenyl (PCB-153), 2,2’,4,4’-tetrabromodiphenyl ether (BDE-

47), and 3,3’,4,4’-tetrabromodiphenyl ether (BDE-77) were obtained from AccuStandard

(New Haven, CT). 13C-PCB-153 was purchased from Wellington Laboratories (Guelph,

ON). Ultra Scientific (North Kingstown, RI) supplied p-terphenyl. Saltwater used in experiments was prepared from ACS Grade NaCl (BDH Chemicals) and deionized water to avoid introduction of dissolved organic carbon, which might alter partitioning. Chemical properties of HOCs used in this study can be found in Table 1.

Polymer pellets used were high-density PE (⌀ 3.7 mm; Rigidex HD6070EA,

Goodfellow Cambridge Ltd.); rigid PVC (ℓ = 4.7 mm; Geon™ Vinyl Rigid Extrusion

6935, PolyOne Corp.); PLA (⌀ 4.8 mm; 100% Natural Virgin, IC3D, LLC); and PHB (⌀

4.9 mm; Mirel DP9002, Metabolic Corp.). Polymer physicochemical characteristics are presented in Table 2.

2.2 Time-to-Steady-State Determination

Batch sorption was used to determine the time-to-steady-state of each polymer/HOC combination with one reaction vessel per time-point. 1L glass amber bottles were filled with 500 mL of autoclaved 35 ppt salt water and spiked with 1mL of 2 μg/mL PCB-153,

BDE-47, and Pyr solution in acetone for a final concentration of 4μg/L. Acetone concentration was kept below 0.1% to minimize solvent effects. Polymer pellets (0.5 g) were added to each bottle, and then mixed on an orbital platform shaker at 100 rpm for 0,

12, 24, 48, 72, 96, 120 and 144 hr. Exposures were performed in duplicate and held at room temperature (19.8 ± 2 ℃).

At each time-point, the aqueous phase (500 mL) was extracted in a 2L separatory funnel using three successive 50 ml aliquots of dichloromethane. Analyte recoveries for the

13 aqueous phase were measured by assessing recoveries of a known amount of d10-Pyr, C-

PCB-153 and BDE-77 added immediately prior to solvent extraction. Solvent extracts were concentrated under high-purity N2 and analyzed by gas chromatography/mass spectrometry (GC/MS) using p-terphenyl as the internal standard. HOC concentrations were corrected for surrogate recoveries before further analysis. The percent removal of each analyte in the water was then plotted versus time.

2.3 HOC Sorption to PE, PVC, PLA, & PHB

For each polymer/individual HOC combination, reaction vessels were set up as above, with final HOC isotherm concentrations ranging from 1.0 – 4.0 μg/L. Each reaction was examined in duplicate at either 48 hrs or 120 hrs, depending on the results of the time-to- steady-state determination, with each reaction held at 20.4 ± 3℃.

At the end of the equilibrium period, plastics were separated from the aqueous phase.

Aqueous phases were extracted and analyzed as described above. HOC accumulation on the glass was modest (average < 4% across all experiments), and so concentration on the polymer beads was calculated by difference. Sorption isotherms were calculated by plotting the concentration of the HOC in the water versus that on the polymer beads.

2.4 Competitive Sorption to PE

Competitive sorption experiments were carried out as described above with PE as the sorbent, except vessels were spiked with 1mL mixture of HOCs (Pyr, PCB-153, BDE-47), wherein each HOC was present in equal concentrations over a 1.0 – 4.0 µg/L range. Each experiment was carried out in duplicate and held at room temperature (18.0 ± 2℃.) After

48hrs, PE was separated from the aqueous phase and the extraction was carried out as described above. Sorption isotherms were calculated by plotting the water HOC concentration versus that on PE at the end of 48 hours.

2.5 GC/MS Parameters

An Agilent 7890A gas chromatograph (GC) and 5975 C mass spectrometer (MS;

Agilent Santa Clara, CA, USA) were used in this study. Ultra-pure grade helium (99.995%:

Airgas) was used as the carrier gas. A J&W DB-5 MS (30 m × 250 μm i.d. × 0.1 μm thickness) was installed. One μL of sample was injected at 260°C by a Gerstel

Multipurpose Sampler in the splitless mode. Initial oven temperature was 40°C held for 2 min, ramping to 250°C at 15°C/min, held for 5 minutes for a total run time of 28 min. The

MS was operated in electron ionization (EI) mode, with an ion source temperature of 230°C and a quadrupole temperature of 150°C. MS was run in a selected ion monitoring (SIM) acquisition.

2.6 Data Analysis

All statistical analyses were performed using R software (version 3.5.1) (2015).

Isotherm models were assessed based on the shapes generated (e.g., linear or exponential) and verified by interpreting r2 values (Table 3) and the small sample-size corrected Akaike

Information Criteria (AICc; Table 4) (Akaike 1974; Burnham & Anderson 2002).

Partitioning coefficients were calculated for each polymer/HOC combination according to two equations: 1) the linear isotherm model based on Henry’s Law (Limousin et al. 2007):

01 = K341 Eq. 1 where 01 is the concentration of HOC sorbed to the polymer, 41 is the concentration of

HOC remaining in the aqueous phase, and K3 is the partitioning coefficient (the slope); and 2) the Freundlich isotherm model, which accounts for surface energy heterogeneity:

78 01 = K5 ∗ 41 Eq. 2 where K5 is the Freundlich partitioning coefficient, and n5 characterizes the energy distribution of sorptive sites on a heterogenous surface (Farrell &Reinhard 1994; Delle Site

2001; Limousin et al. 2002). A Freundlich exponent n5 = 1 indicates equal energies across sites (i.e. homogenous surface), even as surface concentrations of the sorbate increase, reducing the isotherm to the linear approximation (Delle Site 2001; Limousin et al. 2002). n5 < 1 denotes that energy increases with increasing surface concentration and implies

penetration of the solute into the sorbent (absorption), and n5 > 1 relates to energy of sorptive sites decreasing with increasing surface concentration, implying an adsorptive process with limited sites available (‘hole-filling’; Farrell & Reinhard 1994; Delle Site

2001; Limousin et al. 2002).

Differences between the partitioning coefficients of the sorbed HOCs onto the different polymer types were evaluated by two-way analysis of variance (ANOVA). Tukey’s

Honestly Significant Difference (Tukey’s HSD) test was used as a post-hoc analysis.

Partitioning coefficients calculated for each PE/HOC combination were compared to the partitioning coefficients calculated with the HOC mixture. Welch’s two-sample t-test was used to determine if K3 of the single solute system was greater than K3 of the multisolute system.

3. Results & Discussion

3.1 Time-to-Steady-State Determination

The percent removal of HOC from the aqueous phase was plotted versus time sampled

(Figure 1). Steady-state sorption was considered reached sorption stabilized within a 5% range. For most polymer/HOC combination, sorption steady state was reached within 48 hours (Figure 1), except for the PVC/HOC combinations, which required up to 120 hours.

Equilibration times for the HOC isotherm experiments were chosen based on this information. Maximum uptakes of the HOCs for PE, PVC, PHB, and PLA after 144 hours at the initial aqueous concentration of 4 µg/L are presented in Table 3.

Differences seen in the extent of HOC sorption between the different polymers may be due to differences in the physicochemical characteristics of each polymer, e.g. % crystallinity and shape (Table 2). The size and shape of the polymer bead defines the

surface-to-volume ratio and diffusional length scales, which are key in determining the rate of HOC sorption. The data suggest a two-phase sorption process between PVC and HOCs, an initial rapid uptake, followed by a slower secondary association (Figure 1). This may be due to the amorphous PVC existing in a glassy state at room temperature. Glassy amorphous polymers have more condensed and cross-linked polymer chains than rubbery amorphous polymers, and thus have lower diffusivity (Hartmann et al. 2017), which corresponds to slower penetration of the HOCs investigated.

3.2 Modeling

Figures 2 and 3 depicts the fit of the linear isotherm and the Freundlich isotherm model of each polymer and the different HOCs. Table 3 shows the isotherm parameters calculated from each model. All HOCs exhibited stronger partitioning to PE than to the other polymers, with higher logK3 and logK5, and lower n5 values calculated from Eq. 1 and Eq.

2. AICc for both model fits can be found in Table 4. Both models exhibited high r2 (> 0.9), the exception being the PE/BDE-47 data, which were much better modeled by the

Freundlich isotherm (Table 3, bold). Most log K5 exceeded log K3 values, and also exhibited greater standard errors. For simplicity, log K3 were used for statistical analysis.

3.3 HOC Sorption to PE, PVC, PLA, & PHB

Figure 4 compares the log transformed distribution coefficients K3 from the isotherm model for Pyr, PCB-153, and BDE-47 across the four polymer types. The ANOVA table and Tukey’s HSD results are presented in Table 5 & 6. As expected, the various polymer/HOC combinations exhibited differing partitioning, with PE exhibiting the greatest partitioning coefficients across all three HOCs tested. PE used has a higher surface hydrophobicity (Wypych 2012) than the other polymers used in this study, and a high %

crystallinity (Guo et al. 2012). Greater crystallinity reduces sorption of HOCs due to the more condensed and rigid polymer chains in these regions, although at room temperature the remaining amorphous regions of PE are considered rubbery (Table 2). These less organized areas of the plastic structure allow for greater HOC diffusivity (absorption) due to greater free volume between the polymer molecules (Xing & Pignatello 1997; van Noort et al. 2004; Cornelissen et al. 2005; Guo et al. 2012). The Freundlich model fitting of the

PE data also support this idea of greater diffusivity, as all n5 values calculated with PE as the sorbant are less than 1 indicating some level of diffusion into the polymer. This diffusion is most evident for PE/BDE-47 (Table 3, bold).

Interestingly, PVC and PHB exhibited statistically similar log K3 for PCB-153 (p =

0.98) and BDE-47 (p = 0.11). The log K3 of PLA for PCB-153 and BDE-47 were also lower than on PE. This could stem from non-planar PCB-153 and BDE-47 interacting with the functional groups present on these polymers (-Cl for PVC, and -CH3, -O for PHB, PLA;

Table 2) (van Noort et al. 2004). From investigations into sorption onto soil organic matter

(SOM), sorption in “glassy” polymers is by dissolution and hole-filling mechanisms, and so the presence of these functional groups on the polymer backbone can limit the “size” of sorption sites, inhibiting the sorption of large HOCs like PCBs and PBDEs (Xing &

Pignatello 1997; van Noort et al. 2004; Cornelissen et al. 2005). Molecular volume of the three HOCs investigated could also play a role (Table 1). While the amorphous regions of

PHB are rubbery at room temperature, there is a smaller fraction of amorphous regions compared to PE. I anticipated sorption to PHB to be lower based on this, and the additional consideration of polymer structure may encourage sorption of HOCs to appear more along

the lines of the “glassy” polymers, PVC and PLA. This suggests that chemical structure of the polymer sorbate also plays a role in regulating HOC sorption.

If diffusivity and abundance of rubbery domains in polymers regulate HOC sorption

(Xing & Pignatello 1997; van Noort et al. 2004; Cornelissen et al. 2005; Wang & Xing

2007; Teuten et al. 2007; Guo et al. 2012), partitioning of HOCs to the polymers investigated here should favor the most amorphous (low % crystallinity) and rubbery (RT

> Tg) polymers, following the order PE > PHB > PLA > PVC. The data do not follow this pattern of increasing rubbery domains. Hence other factors, including surface hydrophobicity, may contribute. For the polymers utilized in this study, PE has the greatest hydrophobicity based on contact angle with water, followed by PVC, PLA, and PHB. This order more closely matches the overall pattern observed across HOCs: log K3 PE > PVC >

PHB > PLA. Overall, the two bio-based plastics, PHB and PLA exhibited lower affinity for the HOCs investigated in this study compared to the more widely used, petrochemical- derived polymer PE. Partitioning to PVC tracks more closely with PHB and PLA, possibly due to the -Cl surface moieties and the glassy crystalline structure at room temperature.

3.4 Competitive sorption to PE

The linear and Freundlich isotherm model fits for the multi-solute data for PE are shown in Figure 5. Figure 6 shows the single solute isotherm data plotted alongside the multi-solute isotherm data, showing a negative shift in the slope of all isotherms in the multi-solute trial indicating decreased partitioning between the plastic and the surrounding water. The effect was most dramatic for the more hydrophobic PCB-153 and BDE-47, highlighting the competitive effect between these two HOCs. Table 7 shows the isotherm parameters calculated from the linear and Freundlich models for both trials. The

partitioning coefficient calculated using the linear isotherm K3 was then compared to results obtained in the single solute, noncompetitive experiments (Figure 7). Comparing the log transformed K3 from the single sorbate and multi-sorbate experiments directly further demonstrates the effect of competition between the HOCS. Table 8 shows the results of Welch’s two-sample t-test for both experiments. K3 of Pyr onto PE was not statistically different (α = 0.05, H0 = K3BB > K3CB) between the single sorbate and multi- sorbate isotherm trials (p = 0.09). Conversely, both the more hydrophobic PCB-153 and

BDE-47 exhibited highly statistically significant reductions in K3 (p = 0.0002, 0.002) in the multi-sorbate experiments, indicating that competition between HOCs leads to lower individual partitioning.

The multi-solute trial saw a decrease in K3 over the single solute experiments of 2.4%,

19.0%, and 14.6% for Pyr, PCB-153, and BDE-47, respectively, indicating that sorption in the natural environment likely differs from that reported in laboratory single sorbate studies. A competitive index can be calculated from the ratio of 01 in the multi-solute trial to 01 in the single solute trial 01DE/01EE (Bakir et al. 2012). A ratio closer to 1 indicates the absence of competition, and a value closer to zero shows higher competitive behavior

(Table 10). Pyr exhibits no competitive behavior for binding sites (average 01DE⁄01EE =

1.00, and PCB-153 and BDE-47 exhibit only slight competitive behavior

(average. 01DE⁄01EE = 0.89 & 0.94, respectively). Competitive behavior of PCB-153 increased at lower concentrations, but it decreased for BDE-47 at lower concentrations

(Table 10).

Bakir and colleagues (2012) investigated the competitive sorption of phenanthrene

(Phe) and DDT (chemical properties Table 1) to ultra-high molecular weight PE and

unplasticized PVC. They found that in a two-solute system, DDT outcompeted Phe for sorption onto PE. This competition resulted in lower partitioning for both contaminants, but similar overall 01 for DDT in both systems, while 01 for Phe was impacted greatly.

Bakir et al. observed no competition for binding sites for DDT sorption onto PE in the presence of Phe and a strong competitive effect for Phe sorption onto PE, even when Phe was present in greater concentrations. They attributed this difference to the higher affinity

(high hydrophobicity or log K%&) of DDT for PE (Bakir et al. 2012). This K%& based partitioning has also been found in experiments with SOM (Yu & Huang 2005). In the current study, partitioning of Pyr in the multi-solute systems was not impacted by the presence of high K%& PCB-153 and BDE-47, as evidenced by the statistically insignificant difference in log K3 (Figure 6) and the competitive index of 1.00 (Table 10) calculated between the single and multi-solute experiments. This contrasts with what was reported by

Bakir et al. and Yu & Huang. Conversely, Xiao (2004) investigated the competitive sorption between naphthalene (Naph; chemical properties Table 1) and Phe to soils and sediments, finding that although Phe has the higher K%&, Naph outcompeted Phe on SOM.

This is more in line with what was observed in the present study; i.e. Pyr outcompeted the higher K%& PCB-153 and BDE-47 on PE.

The observed differences in competition between solutes could be due to differences in molecular volume in an absorption driven interaction (Table 1). In this present study,

Pyr has a much smaller molecular volume (188.60 Å) than PCB-153 (236.67 Å) or BDE-

47 (235.98 Å). It is possible that Pyr was able to occupy smaller micropores on the surface of the PE bead that were not accessible to the other sorbents. This would explain the lack of competitive pressure on Pyr. PCB-153 and BDE-47 molecules occupy similar volumes

(Table 1) and hence exhibit significant competitive pressures on one another. In the case of Naph and Phe from Xiao et al. (2004), Naph (128.03 Å) occupies a slightly smaller molecular volume than Phe (172.03 Å), leaving smaller binding sites available. In the binary solute experiment of Bakir et al., the much higher affinity (K%&) of DDT for PE likely dominated the interaction between the two solutes and molecular volume did not play a role (256.46 Å vs. 172.03 Å).

Furthermore, PE is a rubbery polymer at RT, and isotherms are predicted to be linear and non-competitive if work in SOM applies (Xing et al. 1996). But it has been shown that sorption to polymers is via a combination of a pore-filling (molecular volume driven) and partitioning (K%& driven) method which fosters competition between solutes (Teuten et al.

2007; Pignatello 1998). This could explain the increased in non-linearity in PE/BDE-47 isotherm in the mixed solute experiment. These two contrasting ideas between partitioning based on hydrophobic interactions and pore-filling adsorption has been discussed elsewhere (Xia and Ball 2000, and enclosed references). It should be noted that I was unable to distinguish individual competitive effects of each HOC on the other due to experimental constraints that allowed for only a single tri-solute experiment to be carried out in tandem with the single solute experiment.

4. Conclusions

Polymers in the environment sorb contaminants to varying degrees, as a function of polymer composition, contaminant properties and the surrounding environmental conditions. Sorption itself will reduce ambient contaminant concentrations. Indeed, polymers are sometimes used in water treatment (Pan et al. 2009), to effectively remove phenolic compounds, organic acids, and aromatic hydrocarbons, reducing general

exposure. However, if sorbed contaminants concentrate on the polymer and are later ingested it has been suggested that exposure may also be increased (Teuten et al. 2009), akin to food chain magnification. Physical degradation of the contaminant/polymer complex may also result in the release of the contaminant over time. As a starting point, an understanding of polymer/contaminant sorption interactions is critical. Again, as shown for

SOM, sorption/desorption from rubbery domains is expected to be faster than glassy domains, but slow desorption still occurs (Xing & Pignatello 1997; Cornelissen et al.

2005). This suggests that bio-based polymers that are mostly crystalline and glassy will release sorbed contaminants only after associated activation energies have been overcome

(60-100 kJ/mol for PAHs and PCBs in whole sediments; Cornelissen et al. 2005). Bio- based polymers such as PHB are expected to breakdown more rapidly in the marine environment than widely used, traditional polymers such as PE and PVC. It is likely that sorbed HOCs will become more bioavailable. PHB degrading bacteria are widespread in the environment (Jendrossek & Handrick 2002) since in situ PHA production is a common carbon storage strategy. These bacteria may also be able to utilize associated HOCs. For instance, several species of the bacterial genus Pseudomonas possess the gene phaZ, which codes for polyhydroxyalkanoate depolymerase, in addition to dehalogenase genes that key for remediating halogenated contaminants such as PCBs and PBDEs (Winsor et al. 2016).

To make bio-based polymers a viable option to replace conventional plastics in all applications, additives will have to be added to some to provide certain desired qualities such as flexibility, durability, reduced flammability, color, etc. (Lambert et al. 2017). These additives, some with toxicological potentials, may later leach into the environment

(Rochman et al. 2014; Koelmans et al. 2014; Bejgarn et al. 2015; Suhrhoff & Scholz-

Böttcher 2016; Hermabessiere et al. 2017; Kwon et al. 2017). Therefore, plastics may represent an important exposure pathway of biota to this often unknown (i.e. trade secret) and constantly evolving suite of chemicals.

Competition for sorptive sites on plastic debris may have substantial implications for assessing the potential of different plastics to sorb and transport HOCs in waterways, impact organisms through ingestion and desorption, and act as sinks for organic contaminants (Yang et al. 2006). Though the goal of this study was not to fully explore mechanistic aspects of competitive sorption onto polymers, it did highlight and reinforce the concept that partitioning is altered for individual constituents present in complex mixtures. Also it underlines that factors beyond K%&, such as chemical structure of the polymer and molecular size of the contaminants of interest, should be considered when evaluating sorption potential to plastic debris. Further studies into competitive sorption should also consider polymer weathering and breakdown, as these processes affect the hydrophobicity of polymers (Suresh et al. 2011) and their surface areas (Feldman 2002), along with the distribution and size of micropores on their surfaces (Feldman 2002).

Additionally, an understanding of the environmental consequences of plastic debris cannot be complete without attention to biofilms. Biofilms (conglomerates of bacteria, algae, protozoans, fungi and other organisms) rapidly colonize newly submerged surfaces also have the potential to change the partitioning of contaminants and release of additives

(Wolfaardt et al. 1994; Headley et al. 1998; Wicke et al. 2007; Wen et al. 2015; Rummel et al. 2017). Exopolymeric substances secreted by biofilms alter sorptive sites. Biofilms are often the target for grazers such as snails (Rummel et al. 2017). Hence film-covered particles may be more likely ingested by selective and passive feeders alike, such as

zooplankton and oysters. If contaminants or leachates accumulate in biofilms, such organisms may be exposed to polymer additives and sorbed HOCs that are not accounted for in existing bioaccumulation/biomagnification models, such as those by Koelmans and colleagues (2016). In conclusion, further investigation is critical into partitioning of environmental contaminants to biofilms in different aquatic environments in order to fully understand the potential effects of plastic debris.

4. Tables

Table 1. Hydrophobic organic contaminants of interest in this study and reference studies

This study: S MW analyte ID name structure log K MV (Å)d %& (μg/L) (g/mol)

Pyra pyrene 4.88 77.0 202.25 188.60

2,2’,4,4’,5,5’- PCB-153b hexachloro- 6.72 0.862 360.86 236.67 biphenyl

2,2’,4,4’- BDE-47c tetrabromo- 6.81 15.0 485.80 235.98 diphenyl ether

Bakir et al. 2012:

4,4’- dichloro- DDT diphenyl- 6.91 5.50 354.48 256.46 trichloroethane

Phe phenanthrene 4.46 1100 178.23 172.03

Xiao 2004:

Naph naphthalene 3.30 3100 128.17 128.03

K%& – octanol/water partitioning coefficient; S – solubility in water at 25°C; MW – molecular weight; MV – molecular volume; aATSDR 2000; bATSDR 1995; cASTDR 2004; dMolinspiration Property Calculation Service.

Table 2. Polymers used in this study and their properties

contact polymer density structure angle w/ Tg (°C) crystalinity ID (g/cm3)a shape water (°)

b -118 to b elliptical HDPE 0.96 135.6 a 58.5% -133 cylinder

83.2 – a a PVC 1.53 a 82 – 87 4 – 10 % rectangu 91.9 lar prism

PHB 1.22 75c 2 – 15e 60 %e elliptical cylinder

d e e PLA 1.24 77 45 – 60 37% ellipsoid

Tg – glass transition temperature; aWypych 2012; bGuo et al. 2012; cKang et al. 2001; dParagkumar et al. 2006; eRoohi et al. 2018

Table 3. Isotherm parameters for single solute sorption onto PE, PVC, PHB, and PLA

Linear modela Freundlich modela

maximum Polymer HOC log K b r2 log K c n d r2* % & & uptake (μg/g)a

PE Pyr 3.8 ± 0.0007 0.98 4.4 ± 0.09 0.81 ± 0.05 0.98 3.58 ± 0.075

PVC Pyr 3.5 ± 0.02 0.98 4.0 ± 0.06 0.82 ± 0.05 0.99 3.92 ± 0.041

PHB Pyr 3.0 ± 0.05 0.97 2.6 ± 0.3 1.1 ± 0.13 0.97 2.81 ± 0.041

PLA Pyr 3.2 ± 0.0006 0.99 3.3 ± 0.06 0.99 ± 0.02 0.99 2.31 ± 0.058

PE PCB-153 5.2 ± 0.02 0.98 5.3 ± 0.2 0.88 ± 0.07 0.98 3.35 ± 0.12

PVC PCB-153 3.8 ± 0.008 0.96 4.4 ± 0.2 0.77 ± 0.07 0.90 3.78 ± 0.25

PHB PCB-153 3.8 ± 0.001 0.97 4.2 ± 0.1 0.83 ± 0.07 0.97 3.46 ± 0.19

PLA PCB-153 3.4 ± 0.002 0.95 4.2 ± 0.1 0.73 ± 0.08 0.96 3.01 ± 0.17

PE BDE-47 4.9 ± 0.0001 0.89 5.8 ± 0.03 0.48 ± 0.02 0.99 3.44 ± 0.10

PVC BDE-47 3.8 ± 0.01 0.99 4.1 ± 0.01 0.90 ± 0.04 0.99 3.88 ± 0.16

PHB BDE-47 3.8 ± 0.01 0.98 4.5 ± 0.04 0.74 ± 0.04 0.99 3.55 ± 0.088

PLA BDE-47 3.0 ± 0.04 0.97 3.0 ± 0.6 0.97 ± 0.09 0.95 2.90 ± 0.51

a ± SE. b Distribution coefficient (L kg-1). c Fruendlich constant (L kg-1). d Fruendlich exponent. *calculated from the linearized Freundlich model [log )* ~ log -.]

Table 4. Small sample-size Akaike Information Criteria (AICc) for isotherm models. polymer HOC model Ka AICc ∆AICc wb

PE Pyr Linear 2 137.76 2.18 0.25 PE Pyr Freundlich 3 135.59 0.00 0.75 PVC Pyr Linear 2 134.88 1.49 0.32 PVC Pyr Freundlich 3 133.39 0.00 0.68 PHB Pyr Linear 2 131.15 0.00 0.94 PHB Pyr Freundlich 3 136.52 5.37 0.06 PLA Pyr Linear 2 104.60 0.00 0.65 PLA Pyr Freundlich 3 105.81 1.21 0.35

PE PCB-153 Linear 2 137.71 0.00 0.89 PE PCB-153 Freundlich 3 141.92 4.21 0.11 PVC PCB-153 Linear 2 142.27 0.00 0.62 PVC PCB-153 Freundlich 3 143.27 1.00 0.38 PHB PCB-153 Linear 2 138.75 0.00 0.65 PHB PCB-153 Freundlich 3 140.03 1.28 0.35 PLA PCB-153 Linear 2 141.19 1.71 0.30 PLA PCB-153 Freundlich 3 139.48 0.00 0.70

PE BDE-47 Linear 2 157.20 27.18 0.00 PE BDE-47 Freundlich 3 130.02 0.00 1.00 PVC BDE-47 Linear 2 125.79 0.00 0.64 PVC BDE-47 Freundlich 3 126.90 1.11 0.36 PHB BDE-47 Linear 2 138.73 6.45 0.04 PHB BDE-47 Freundlich 3 132.28 0.00 0.96 PLA BDE-47 Linear 2 132.12 0.00 0.94 PLA BDE-47 Freundlich 3 137.72 5.60 0.06 a number of parameters of the model; b weight of evidence in favor of the model

Table 5. Two-way analysis of variance data for all HOC/polymer combinations.

HOC Df sum of sq. mean sq. F p (>F) Pyr factor 3 0.8543 0.28475 204.6 7.84E-05 * residuals 4 0.0056 0.00139

PCB-153 factor 3 3.624 1.208 5240 1.21E-07 * residuals 4 0.001 0.002

BDE-47 factor 3 3.718 1.2392 1128 2.61E-06 * residuals 4 0.004 0.0011

* indicates mean !" statistically different at α = 0.05

Table 6. Results of Tukey HSD analysis of ANOVA results. polymer p comparisons Pyr PCB-153 BDE-47

PE-PVC 2.60E-03 1.22E-13 3.14E-05

PE-PHB 8.92E-05 1.17E-13 1.82E-05

PE-PLA 2.98E-04 3.97E-14 4.22E-08

PVC-PHB 4.62E-04 0.976 * 0.114 *

PVC-PLA 7.87E-03 1.04E-04 5.76E-05

PLA-PHB 5.72E-03 1.10E-04 8.78E-05

* indicates mean !" statistically similar at α = 0.05

Table 7. Isotherm parameters for single solute and multi-solute experiments on PE.

a a System HOC Linear model Freundlich model b 2 c d 2 log K' r log K) n) r *

Pyr 3.9 ± 0.0007 0.98 4.4 ± 0.09 0.81 ± 0.05 0.98

Single PCB-153 5.2 ± 0.02 0.98 5.3 ± 0.2 0.88 ± 0.07 0.98 Solute BDE-47 4.9 ± 0.0001 0.89 5.8 ± 0.03 0.48 ± 0.02 0.99

Pyr 3.8 ± 0.03 0.96 4.0 ± 0.4 0.83 ± 0.09 0.96 Multi- PCB-153 4.2 ± 0.002 0.97 4.4 ± 0.1 0.92 ± 0.08 0.94 Solute BDE-47 4.2 ± 0.01 0.91 5.5 ± 0.07 0.46 ± 0.03 0.98 a ± SE. b Distribution coefficient (L kg-1). c Fruendlich constant (L kg-1). d Fruendlich exponent. *calculated from the linearized Freundlich model [log ,- ~ log /0]

Table 8. Welch's two-sample t-test comparing mean log K' of PE of single solute and multi-solute systems

a b log K'22 log K'32 t-stat Df p

Pyr 3.86 3.76 3.52 1.0 0.0880 PCB-153 5.18 4.20 49.4 2.0 0.0002 ** BDE-47 4.88 4.17 58.6 2.0 0.0015 **

* indicates a significant reduction in mean K' between the asingle solute and bmulti-solute HOC mixture.

Table 10. Competitive indices for the sorption of Pyr, PCB-153, and BDE-47 onto PE for four concentration ranges.

,045 a b 6 ,045 ,055 ,055 Pyr 3.34 ± 0.04 3.34 ± 0.04 1.00 2.57 ± 0.07 2.53 ± 0.02 1.02

1.69 ± 0.04 1.70 ± 0.03 0.99

0.85 ± 0.01 0.85 ± 0.01 1.00

mean 1.00

PCB-153 3.54 ± 0.06 3.89 ± 0.06 0.91 2.49 ± 0.02 2.87 ± 0.06 0.87

1.73 ± 0.01 1.94 ± 0.00 0.89

0.83 ± 0.01 0.93 ± 0.02 0.89

mean 0.89

BDE-47 3.56 ± 0.02 3.93 ± 0.08 0.91 2.70 ± 0.01 2.91 ± 0.07 0.93

1.89 ± 0.04 1.98 ± 0.04 0.95

0.96 ± 0.00 0.97 ± 0.04 0.99

mean 0.94

a HOC concentration on PE in multi-solute system; b HOC concentration on PE in single solute system

5. Figures

Pyrene PCB−153

100% 100% 90%

80% 90%

70% 80% 60% Sorption 50% Sorption 70%

40% ● PE PHB 60% ● PE PHB 30% PVC PLA PVC PLA 0% 0%

0 24 48 72 96 120 144 0 24 48 72 96 120 144 Time (hr) Time (hr) BDE−47

100%

90%

80%

Sorption 70%

60% ● PE PHB PVC PLA 0%

0 24 48 72 96 120 144 Time (hr)

Figure 21. Time to Steady State determination. Error bars indicate ± 5%.

PE: Pyr PE: PCB−153 PE: BDE−47

4000 4000 4000

3000 3000 3000

2000 2000 2000 (ng/g) (ng/g) (ng/g) c c c Q Q Q

1000 1000 1000

0 0 0

0 100 200 300 400 500 0 5 10 15 20 25 0 10 20 30 40 50 Ce(ng/L) Ce(ng/L) Ce(ng/L)

PVC: Pyr PVC: PCB−153 PVC: BDE−47

4000 4000 4000

3000 3000 3000

2000 2000 2000 (ng/g) (ng/g) (ng/g) c c c Q Q Q

1000 1000 1000

0 0 0

0 200 400 600 800 1000 0 100 200 300 400 500 600 0 100 200 300 400 500 Ce(ng/L) Ce(ng/L) Ce(ng/L) Figure 2. Linear (solid line) and Freundlich (dashed line) isotherm models fit to the single solute data for petrochemical polymers.

PHA: Pyr PHA: PCB−153 PHA: BDE−47

4000 4000 4000

3000 3000 3000

2000 2000 2000 (ng/g) (ng/g) (ng/g) c c c Q Q Q

1000 1000 1000

0 0 0

0 500 1000 1500 2000 0 100 200 300 400 500 0 100 200 300 400 500 600 Ce(ng/L) Ce(ng/L) Ce(ng/L) PLA: Pyr PLA: PCB−153 PLA: BDE−47

4000 4000 4000

3000 3000 3000

2000 2000 2000 (ng/g) (ng/g) (ng/g) c c e Q Q Q

1000 1000 1000

0 0 0

0 500 1000 1500 2000 0 200 400 600 800 1000 0 500 1000 1500 2000 2500 Ce(ng/L) Ce(ng/L) Ce(ng/L)

Figure 3. Linear (solid line) and Freundlich (dashed line) isotherm models fit to the single solute data for bio-based polymers.

Pyr PCB−153 BDE−47 5.50 5.50 5.50 5.00 5.00 5.00 4.50 4.50 4.50 A A A 4.00 4.00 4.00 A 3.50 3.50 3.50 d d d 3.00 3.00 3.00 logK logK 2.50 2.50 logK 2.50

2.00 2.00 2.00

0.00 0.00 0.00 PE PE PE PLA PLA PVC PHB PVC PHB PLA PVC PHB

Figure 4. Comparison of log transformed K" across all polymers and HOCs investigated. Error bars indicate 95% confidence interval. A indicates sstatisticatatistically similar log K" at the α = 0.05 level.

PE−multi: Pyr PE−multi: PCB−153 PE−multi: BDE−47

4000 4000 4000

3000 3000 3000

2000 2000 2000 (ng/g) (ng/g) (ng/g) c c c Q Q Q

1000 1000 1000

0 0 0

0 100 200 300 400 500 600 0 50 100 150 200 250 0 50 100 150 200 250 Ce(ng/L) Ce(ng/L) Ce(ng/L) Figure 5. Linear (solid line) and Freundlich (da(dashedshed line) isotherm model fits to the multi-solute data. PE: Pyr PE: PCB−153 PE: BDE−47 4000 4000 ● 4000 ● ● 3000 3000 ● 3000 ● ●

2000 2000 2000

(ng/g) (ng/g) ● (ng/g) ● c ● c c Q Q Q

1000 1000 ● 1000 ● ● ● Single Multi. 0 ● 0 ● 0 ● 0 100 200 300 400 500 600 0 50 100 150 200 250 0 25 50 75 100 125 150 175 200 225 Ce(ng/L) Ce(ng/L) Ce(ng/L) Figure 6. Single solute (circle) and multi-solute (inverted triangle) isotherm data.

Pyr PCB−153 BDE−47 5.50 5.50 5.50 5.00 5.00 5.00 4.50 4.50 4.50 4.00 4.00 * 4.00 * 3.50 3.50 3.50 d d d 3.00 3.00 3.00 logK logK 2.50 2.50 logK 2.50

2.00 2.00 2.00

0.00 0.00 0.00 Single Multi. Single Multi. Single Multi.

Figure 7. Comparison of distribution coefficients between single solutesolute and multi-solute systems for PE. ErrErroror bars indicate 95% confidence intervals. * indicates a statistically significant decreasedecrease in log K%.

6. References

Aeschelmann, F. & Carus, M., 2016. Bio-based Building Blocks and Polymers. Global Capacities and Trends 2016 – 2021.

Akaike, H., 1974. A new look at the statistical model identification. IEEE Transactions on Automatic Control, 19(6), pp.716–723. Available at: http://ieeexplore.ieee.org/document/1100705/.

Andrady, A.L., 2015. Persistence of Plastic Litter in the Oceans. In M. Bergmann, L. Gutow, & M. Klages, eds. Marine Anthropogenic Litter. Cham: Springer International Publishing, pp. 57–72. Available at: http://link.springer.com/book/10.1007%2F978-3-319-16510-3.

ASTDR, 2000. Toxicological Profile for Polychlorinated Biphenyls (PCBs). Agency for Toxic Substances and Disease Registry, (November), pp.1–948. Available at: http://www.atsdr.cdc.gov/toxprofiles/tp.asp?id=142&tid=26.

ASTDR, 2004. Toxicological Profile for Polybrominated Diphenyl Ethers (PBDEs). Agency for Toxic Substances and Disease Registry, (September), pp.1–599. Available at: https://www.atsdr.cdc.gov/ToxProfiles/tp.asp?id=901&tid=183#bookmark16.

ASTDR, 1996. Toxicological Profile for Polycyclic Aromatic Hydrocarbons (PAHs). Agency for Toxic Substances and Disease Registry, (August), pp.1–487. Available at: https://www.atsdr.cdc.gov/ToxProfiles/tp.asp?id=122&tid=25.

Bakir, A., Rowland, S.J. & Thompson, R.C., 2012. Competitive sorption of persistent organic pollutants onto microplastics in the marine environment. Marine Pollution Bulletin, 64(12), pp.2782–2789. Available at: http://dx.doi.org/10.1016/j.marpolbul.2012.09.010.

Bejgarn, S. et al., 2015. Toxicity of leachate from weathering plastics: An exploratory screening study with Nitocra spinipes. Chemosphere, 132, pp.114–119. Available at: http://dx.doi.org/10.1016/j.chemosphere.2015.03.010.

Burnham, K.P. & Anderson, D.R., 2002. Model Selection and Multimodel Inference: A Practical Information-Theoretic Approach (2nd ed), Available at: http://linkinghub.elsevier.com/retrieve/pii/S0304380003004526.

Cornelissen, G. et al., 2005. Extensive sorption of organic compounds to black carbon, coal, and kerogen in sediments and soils: Mechanisms and consequences for distribution, bioaccumulation, and biodegradation. Environmental Science & Technology, 39(18), pp.6881–6895. Available at: http://pubs.acs.org/doi/abs/10.1021/es050191b.

Delle Site, A., 2001. Factors affecting sorption of organic compounds in natural sorbent/water systems and sorption coefficients for selected pollutants. A review. Journal of Physical and Chemical Reference Data, 30(1), pp.187–439. Available at: http://scitation.aip.org/content/aip/journal/jpcrd/30/1/10.1063/1.1347984 [Accessed July 6, 2016].

Diepens, N.J. & Koelmans, A.A., 2018. Accumulation of plastic debris and associated contaminants in aquatic food webs. Environmental Science & Technology, 52(15), pp.8510–8520. Available at: http://pubs.acs.org/doi/10.1021/acs.est.8b02515.

Farrell, J. & Reinhard, M., 1994. Desorption of halogenated organics from model solids, sediments, and soil under unsaturated conditions. 1. Isotherms. Environmental Science & Technology, 28(1), pp.53–62. Available at: http://pubs.acs.org/doi/abs/10.1021/es00050a009.

Feldman, D., 2002. Polymer weathering: Photo-oxidation. Journal of Polymers and the Environment, 10(4), pp.163–173.

Gouin, T. et al., 2011. A thermodynamic approach for assessing the environmental exposure of chemicals absorbed to microplastic. Environmental Science & Technology, 45(4), pp.1466–1472. Available at: http://www.ncbi.nlm.nih.gov/pubmed/21268630.

Guo, X. et al., 2012. Sorption of four hydrophobic organic compounds by three chemically distinct polymers: Role of chemical and physical composition. Environmental Science & Technology, 46(13), pp.7252–7259. Available at: http://pubs.acs.org/doi/10.1021/es301386z.

Hartmann, N.B. et al., 2017. Microplastics as vectors for environmental contaminants: Exploring sorption, desorption, and transfer to biota. Integrated Environmental Assessment and Management, 13(3), pp.488–493. Available at: http://doi.wiley.com/10.1002/ieam.1904.

Havens, K.J. et al., 2015. Fishing Gear with Degradable Component, pp.12–15.

Headley, J. V. et al., 1998. Rates of sorption and partitioning of contaminants in river biofilm. Environmental Science and Technology, 32(306), pp.3968–3973. Available at: http://dx.doi.org/10.1021/es980499l [Accessed May 20, 2016].

Hermabessiere, L. et al., 2017. Occurrence and effects of plastic additives on marine environments and organisms: A review. Chemosphere, 182, pp.781–793. Available at: https://linkinghub.elsevier.com/retrieve/pii/S0045653517308007.

Houghton, J.I. & Quarmby, J., 1999. Biopolymers in wastewater treatment. Current Opinion in Biotechnology, 10(3), pp.259–262. Available at: http://linkinghub.elsevier.com/retrieve/pii/S0958166999800457.

Humel, S. et al., 2017. Enhanced accessibility of polycyclic aromatic hydrocarbons (PAHs) and heterocyclic PAHs in industrially contaminated soil after passive dosing of a competitive sorbate. Environmental Science & Technology, 51(14), pp.8017– 8026. Available at: http://pubs.acs.org/doi/10.1021/acs.est.7b01198.

InnProBio, 2016. Sustainability of bio-based products, Available at: http://innprobio.innovation-procurement.org/bio-based-products- services/factsheets/.

Jambeck, J.R. et al., 2015. Plastic waste inputs from land into the ocean. Science, 347(6223), pp.768–771. Available at: http://www.sciencemag.org/cgi/doi/10.1126/science.1260352.

Jendrossek, D. & Handrick, R.R., 2002. Microbial Degradation of Polyhydroxyalkanoates S. N. Singh, ed. Annual Review of Microbiology, 56(1), pp.403–432. Available at: http://link.springer.com/10.1007/978-3-642-23789-8 [Accessed May 24, 2016].

Weber Jr., W. J., Kim, S. H., & Johnson, M. D. (2002). Distributed reactivity model for sorption by soils and sediments. 15. High-concentration co-contaminant effects on phenanthrene sorption and desorption. Environmental Science & Technology, 36(16), 3625–3634. https://doi.org/10.1021/es020557+

Kang, I.-K. et al., 2001. Surface modification of polyhydroxyalkanoate films and their interaction with human fibroblasts. International Journal of Biological Macromolecules, 28(3), pp.205–212.

Kleinteich, J. et al., 2018. Microplastics reduce short-term effects of environmental contaminants. Part II: Polyethylene particles decrease the effect of polycyclic aromatic hydrocarbons on microorganisms. International Journal of Environmental Research and Public Health, 15(2), p.287. Available at: http://www.mdpi.com/1660-4601/15/2/287.

Koelmans, A.A. et al., 2016. Microplastic as a vector for chemicals in the aquatic environment: Critical review and model-supported reinterpretation of empirical studies. Environmental Science & Technology, 50(7), pp.3315–26. Available at: http://apps.webofknowledge.com/full_record.do?product=UA&search_mode=Gener alSearch&qid=1&SID=3DMbOj3EfmjCdudRiMU&page=1&doc=6 [Accessed May 3, 2016].

Koelmans, A. A., Besseling, E. & Foekema, E.M., 2014. Leaching of plastic additives to marine organisms. Environmental Pollution, 187, pp.49–54. Available at: http://dx.doi.org/10.1016/j.envpol.2013.12.013.

Kwon, J.-H. et al., 2017. Microplastics as a vector of hydrophobic contaminants: Importance of hydrophobic additives. Integrated Environmental Assessment and

Management, 13(3), pp.494–499. Available at: http://doi.wiley.com/10.1002/ieam.1906.

Lambert, S. & Wagner, M., 2017. Environmental performance of bio-based and biodegradable plastics: the road ahead. Chemical Society Reviews, 46(22), pp.6855– 6871. Available at: http://dx.doi.org/10.1039/C7CS00149E.

Li, J. & Werth, C.J., 2001. Evaluating competitive sorption mechanisms of volatile organic compounds in soils and sediments using polymers and zeolites. Environmental Science & Technology, 35(3), pp.568–574. Available at: http://pubs.acs.org/doi/abs/10.1021/es001366e.

Limousin, G. et al., 2007. Sorption isotherms: A review on physical bases, modeling and measurement. Applied Geochemistry, 22(2), pp.249–275. Available at: http://linkinghub.elsevier.com/retrieve/pii/S0883292706002629.

Lohmann, R., 2017. Microplastics are not important for the cycling and bioaccumulation of organic pollutants in the oceans-but should microplastics be considered POPs themselves? Integrated Environmental Assessment and Management, 13(3), pp.460– 465. Available at: http://doi.wiley.com/10.1002/ieam.1914.

Mato, Y. et al., 2001. Plastic resin pellets as a transport medium for toxic chemicals in the marine environment. Environmental Science & Technology, 35(2), pp.318–324. Available at: http://dx.doi.org/10.1021/es0010498 [Accessed October 29, 2015].

Matsuzawa, Y. et al., 2010. Removal of hydrophobic organic contaminants from aqueous solutions by sorption onto biodegradable polyesters. Journal of Water Resource and Protection, 02(03), pp.214–221. Available at: http://www.scirp.org/journal/PaperInformation.aspx?PaperID=1492&#abstract [Accessed May 22, 2016].

McDevitt, J.P. et al., 2017. Addressing the issue of microplastics in the wake of the Microbead-Free Waters Act—A new standard can facilitate improved policy. Environmental Science & Technology, 51(12), pp.6611–6617. Available at: http://pubs.acs.org/doi/abs/10.1021/acs.est.6b05812 [Accessed August 28, 2017].

McGinley, P.M., Katz, L.E. & Weber, W.J., 1993. A distributed reactivity model for sorption by soils and sediments. 2. Multicomponent systems and competitive effects. Environmental Science & Technology, 27(8), pp.1524–1531. Available at: http://pubs.acs.org/doi/abs/10.1021/es00045a006.

Molinspiration Property Calculation Service. Available at: http://www.molinspiration.com/ [Accessed August 10, 2018].

Pan, B. et al., 2009. Development of polymeric and polymer-based hybrid adsorbents for pollutants removal from waters. Chemical Engineering Journal, 151(1–3), pp.19– 29. Available at: http://linkinghub.elsevier.com/retrieve/pii/S1385894709001703.

Paragkumar N, T., Edith, D. & Jean-Luc, S., 2006. Surface characteristics of PLA and PLGA films. Applied Surface Science, 253(5), pp.2758–2764. Available at: http://linkinghub.elsevier.com/retrieve/pii/S0169433206007641 [Accessed July 14, 2016].

Rochman, C.M. et al., 2013. Long-term field measurement of sorption of organic contaminants to five types of plastic pellets: Implications for plastic marine debris. Environmental Science & Technology, 47(3), pp.1646–1654. Available at: http://pubs.acs.org/doi/abs/10.1021/es303700s.

Rochman, C.M. et al., 2014. Polybrominated diphenyl ethers (PBDEs) in fish tissue may be an indicator of plastic contamination in marine habitats. Science of The Total Environment, 476–477, pp.622–633. Available at: http://dx.doi.org/10.1016/j.scitotenv.2014.01.058.

Rochman, C.M. et al., 2013. Polystyrene plastic: A source and sink for polycyclic aromatic hydrocarbons in the marine environment. Environmental Science and Technology, 47(24), pp.13976–13984. Available at: http://pubs.acs.org/doi/10.1021/es403605f.

Roohi, Zaheer, M.R. & Kuddus, M., 2018. PHB (poly-β-hydroxybutyrate) and its enzymatic degradation. Polymers for Advanced Technologies, 29(1), pp.30–40. Available at: http://doi.wiley.com/10.1002/pat.4126.

Rummel, C.D. et al., 2017. Impacts of biofilm formation on the fate and potential effects of microplastic in the aquatic environment. Environmental Science & Technology Letters, 4(7), pp.258–267. Available at: http://pubs.acs.org/doi/abs/10.1021/acs.estlett.7b00164 [Accessed August 28, 2017].

Sander, M. & Pignatello, J.J., 2005. Characterization of charcoal adsorption sites for aromatic compounds: Insights drawn from single-solute and bi-solute competitive experiments. Environmental Science and Technology, 39(6), pp.1606–1615.

Shen, X. et al., 2014. Displacement and competitive sorption of organic pollutants on multiwalled carbon nanotubes. Environmental Science and Pollution Research, 21(20), pp.11979–11986. Available at: http://link.springer.com/10.1007/s11356- 014-3115-1.

Suhrhoff, T.J. & Scholz-Böttcher, B.M., 2016. Qualitative impact of salinity, UV radiation and turbulence on leaching of organic plastic additives from four common plastics — A lab experiment. Marine Pollution Bulletin, 102(1), pp.84–94. Available at:

http://apps.webofknowledge.com/full_record.do?product=UA&search_mode=Gener alSearch&qid=1&SID=3DMbOj3EfmjCdudRiMU&page=2&doc=70 [Accessed April 4, 2016].

Suresh, B. et al., 2011. Mechanical and surface properties of low-density polyethylene film modified by photo-oxidation. Polymer Journal, 43(4), pp.398–406. Available at: http://dx.doi.org/10.1038/pj.2010.147.

Teuten, E.L. et al., 2007. Potential for plastics to transport hydrophobic contaminants. Environmental Science & Technology, 41(22), pp.7759–7764. Available at: http://pubs.acs.org/doi/abs/10.1021/es071737s.

Teuten, E.L. et al., 2009. Transport and release of chemicals from plastics to the environment and to wildlife. Philosophical Transactions of the Royal Society B: Biological Sciences, 364(1526), pp.2027–2045. Available at: http://rstb.royalsocietypublishing.org/content/364/1526/2027.short.

Valderrama, C. et al., 2010. Evaluating binary sorption of phenol/aniline fromaqueous solutions onto granular activated carbon and hypercrosslinked polymeric resin (MN200). Water, Air, and Soil Pollution, 210(1–4), pp.421–434. Available at: http://link.springer.com/10.1007/s11270-009-0266-7.

Van Noort, P.C.M., Jonker, M.T.O. & Koelmans, A.A., 2004. Modeling maximum adsorption capacities of soot and soot-like materials for PAHs and PCBs. Environmental Science and Technology, 38(12), pp.3305–3309.

Van, A. et al., 2012. Persistent organic pollutants in plastic marine debris found on beaches in San Diego, California. CHEMOSPHERE, 86(3), pp.258–263. Available at: http://linkinghub.elsevier.com/retrieve/pii/S0045653511010836.

Wang, X. et al., 2007. Sorption of organic contaminants by biopolymers: Role of polarity, structure and domain spatial arrangement. Chemosphere, 66(8), pp.1476– 1484. Available at: http://linkinghub.elsevier.com/retrieve/pii/S0045653506011799.

Wang, X., Sato, T. & Xing, B., 2006. Competitive sorption of pyrene on wood chars. Environmental Science and Technology, 40(10), pp.3267–3272.

Wang, X., Tao, S. & Xing, B., 2009. Sorption and competition of aromatic compounds and humic acid on multiwalled carbon nanotubes. Environmental Science & Technology, 43(16), pp.6214–6219. Available at: http://pubs.acs.org/doi/abs/10.1021/es901062t.

Wang, X. & Xing, B., 2007. Importance of structural makeup of biopolymers for organic contaminant sorption. Environmental Science and Technology, 41(10), pp.3559– 3565.

Wen, G. et al., 2015. Biomig—A method to evaluate the potential release of compounds from and the formation of biofilms on polymeric materials in contact with drinking water. Environmental Science & Technology, 49(19), pp.11659–11669. Available at: http://pubs.acs.org/doi/10.1021/acs.est.5b02539.

Wicke, D., Böckelmann, U. & Reemtsma, T., 2007. Experimental and modeling approach to study sorption of dissolved hydrophobic organic contaminants to microbial biofilms. Water Research, 41(10), pp.2202–2210.

Winsor, G.L. et al., 2016. Enhanced annotations and features for comparing thousands of Pseudomonas genomes in the Pseudomonas genome database. Nucleic Acids Research, 44(D1), pp.D646–D653. Available at: https://academic.oup.com/nar/article-lookup/doi/10.1093/nar/gkv1227.

Wolfaardt, G.M. et al., 1994. Microbial exopolymers provide a mechanism for bioaccumulation of contaminants. Microbial Ecology, 27(3), pp.279–291. Available at: http://www.ncbi.nlm.nih.gov/pubmed/24190341 [Accessed April 9, 2016].

Wypych, G., 2012. Handbok of Polymers, Toronto: ChemTec Publishing.

Xia, G. & Ball, W.P., 2000. Polanyi-based models for the competitive sorption of low- polarity organic contaminants on a natural sorbent. Environmental Science & Technology, 34(7), pp.1246–1253. Available at: http://pubs.acs.org/doi/abs/10.1021/es9812453.

Xing, B. & Pignatello, J.J., 1997. Dual-mode sorption of low-polarity compounds in glassy poly(vinyl chloride) and soil organic matter. Environmental Science & Technology, 31(3), pp.792–799. Available at: http://pubs.acs.org/doi/abs/10.1021/es960481f.

Xing, B., Pignatello, J.J. & Gigliotti, B., 1996. Competitive sorption between atrazine and other organic compounds in soils and model sorbents. Environmental Science & Technology, 30(8), pp.2432–2440. Available at: http://pubs.acs.org/doi/abs/10.1021/es950350z.

Yang, K. et al., 2006. Competitive sorption of pyrene, phenanthrene, and naphthalene on multiwalled carbon nanotubes. Environmental Science and Technology, 40(18), pp.5804–5810. Available at: http://pubs.acs.org/doi/abs/10.1021/es061081n.

Yu, Z. & Huang, W., 2005. Competitive sorption between 17α-ethinyl estradiol and naphthalene/phenanthrene by sediments. Environmental Science & Technology, 39(13), pp.4878–4885. Available at: http://pubs.acs.org/doi/abs/10.1021/es048558k.

Zhao, D. et al., 2002. Application of the dual-mode model for predicting competitive sorption equilibria and rates of polycyclic aromatic hydrocarbons in estuarine sediment suspensions. Environmental Toxicology and Chemistry, 21(11), pp.2276– 2282. Available at: http://doi.wiley.com/10.1002/etc.5620211104.

Characterizing the Microbial Diversity of Estuarine Biofilms on Microplastics

1. Introduction

Plastic marine debris (PMD) is an important global pollutant. Plastics have been documented in the oceans since the 1970s (Carpenter & Smith 1972), only 20 years after the advent of disposable plastic commodities. PMD entry into the oceans continues to grow at a staggering rate, with 4.8 - 12.7 million tonnes available to enter the oceans every year (Jambeck et al. 2015). When plastic debris enters the environment, physical processes (e.g. wave action, abrasion, UV exposure) cause larger pieces to fragment, eventually forming microplastics (MPs; < 5 mm in size). In addition to impacts caused by ingestion of MPs (Sigler et al. 2014), exposure to sorbed chemicals (Mato et al. 2001;

Teuten et al. 2007) and leached additives from microplastic debris (Hermabessiere et al.

2017; Kwon et al. 2017) may occur. Recently, the scientific community has begun to investigate the microbial communities that colonize on plastic debris in aquatic environments (Harrison et al. 2011). Marine-borne plastics may be more persistent than natural polymeric debris (e.g. cellulose, lignin and chitin) common in coastal and open- ocean systems, potentially allowing for greater transport and distribution of associated colonizing microbes. Several recent studies have focused on the potential for PMD to carry pathogenic and invasive species long distances (Zettler et al. 2013; Oberbeckmann et al. 2014, 2016; Debroas et al. 2017; Frère et al. 2018).

Recent efforts have been made to identify a “core microbiome” that colonize on plastic surfaces. Newly submerged solid surfaces rapidly develop a coating of proteins and polysaccharides from the surrounding environment (Jain & Bholse 2010), facilitating the settlement of microbes, commonly referred to as a biofilm. This biofilm can have a range of consequences, most notably altering the buoyancy of the plastic (Ye & Andrady

1991; Lobelle & Cunliffe 2011) and thus the distribution of PMD throughout the water column. Biofilms may also affect the degradation of PMD, by either causing the plastic to sink below the photic zone or absorbing UV radiation at the polymer surface, restricting photo-degradation (Andrady 2015). Alternatively, biofilm communities may harbor plastic degrading members (Sudhakar et al. 2007; Artham et al. 2009; Oberbeckmann et al. 2015, Debroas et al. 2017), contributing to the biodegradation and remineralization of plastic debris.

Zettler et al (2013) coined the term “plastisphere” in reference to the novel microbial communities found on plastic debris collected from the North Atlantic gyre. Since then, evidence has suggested that PMD biofilm communities may differ from both free-living bacterial community (FL) present in surrounding waters (Zettler et al. 2013;

Oberbeckmann et al. 2014; Debroas et al. 2017), but also those on natural particle- associated communities (PA) (McCormick et al. 2014; Dussud et al. 2018a). However, others have disagreed (e.g. Oberbeckmann et al. 2016, Frère et al. 2018), finding no distinct communities on PMD. Indeed, as biofilms age, community compositions may coalesce. To explore this, additional studies focusing on initial periods of biofilm formation (2 – 6 weeks) on deployed PMD (Oberbeckmann et al. 2018; Ogonowski et al.

2018; Dussud et al. 2018b) have been initiated because variances in biofilm communities are thought to be more pronounced in early colonizers.

Differences in substrate characteristics have also been considered with respect to novel plastic biofilm communities. Surface characteristics such as polymer hydrophobicity (Zhang et al. 2015) and roughness (Sweet et al. 2011) may play an important role in initial adsorption of proteins and the subsequent settlement of

colonizing microbes. This initial layer in turn influences secondary colonizers and may lead to differentiation in communities between polymer types as the biofilm matures and stabilizes (Lorite et al. 2011). Natural polymer substrates, such as cellulose and wood, have also recently been investigated alongside PMD, to attempt to distinguish synthetic and natural debris communities (Oberbeckmann et al. 2018; Ogonowski et al. 2018).

To our knowledge, only one study (Dussud et al. 2018) compared biofilm communities forming on a bio-based, degradable plastic compared to conventional plastics. In general, bio-based plastics are made of building blocks derived from renewable, non-fossilized biomass. These bio-based polymers are growing in market share (Aeschelmann & Carus 2016) and may constitute a greater percentage of future

PMD.

Dussud (2018) investigated the stages of biofilm colonization of four polymers; petrochemical-based low-density polyethylene film (LDPE), an LDPE film with a pro- oxidant additive (OXO-PE), an aged OXO-PE film, and the bio-based polymer poly(3- hydroxybutyrate-co-3-hydroxyvalerate) (PHBV), which is a biodegradable polyhydroxyalkanoate (PHA) polyester. PHAs are produced in situ by several bacterial taxa allowing intracellular storage of carbon during periods of excess carbon when growth is limited by another nutrient (e.g. nitrogen). PAHs have been shown to be vulnerable to hydrolysis in the natural environment (Jendrossek & Handrick 2002).

Traditional polymers with pro-oxidative additives are designed to breakdown by abiotic processes (e.g. heat, UV exposure) at an accelerated rate compared to their non-OXO counterparts. Note that this only breaks down the polymer physically and does not directly result in remineralization of the petrochemical polymer base. Dussud et al.

(2018) found that early biofilm communities were similar across substrates, but that

LDPE and OXO-PE formed distinct communities from each other and the other polymers over time. The aged-OXO-PE and PHBV communities were much more similar to each other, but these communities continued to diverge over the course of the 6-week experiment. The authors suggested that this may be due to differences in surface properties (i.e. roughness) and colonization by degrading bacteria (Dussud et al. 2018).

To further study the progression of biofilm development on polymer, the current study uses a next-generation sequence analysis of 16S ribosomal RNA (rRNA) genes to investigate the early estuarine microbial colonizers over a four-week period on five different polymer types: two conventional plastics (high-density polyethylene, PE and polyvinyl chloride PVC), two bio-based polymers (poly-3-hydroxybutyrate PHB and polylactic acid PLA), and the natural polymer chitin. Interestingly, despite high prevalence of use in the marine environment, no studies to date have investigated the microbial colonization of PVC in the natural marine environment. This may be due to the absences of PVC in sea surface collections due to its negative buoyancy (specific gravity

1.53 g/cm3). By focusing on the early colonization period, we aimed to capture subtle differences in the biofilm communities. The biofilms investigated in this study include three different stages of development: early colonization, exponential growth, and the pseudo-stable stage. The characteristics of these stages includes attachment of proteins and adaptation by a wide variety of bacterioplankton (0 – 12 days) followed by exponential accumulation of bacteria and eukarya (12 – 17 days), then a stable state characterized by an equilibrium between dispersal and accumulation of colonizing species (17 days on; Fisher et al. 2014). Dussud et al. also distinguished phases of

biofilm growth through flow cytometry: primo-colonization during the first 7 days, growth between days 7 and 22, and stabilization after 22 days of immersion.

The following questions were formulated: 1) Do early differences in biofilm communities develop between different substrates due to differences in substrate properties (i.e. hydrophobicity and surface roughness), and if so, do these differences resolve over time?; 2) Do biofilm communities on a naturally occurring polymer (chitin) differ from synthetic polymer?; 3) Do the biofilm communities forming on bio-based plastics differ from those on petrochemical plastics?; and 4) Is there an increase of known plastic degrading or hydrocarbonoclastic bacteria and pathogenic bacteria on deployed microplastics?

2. Materials & Methods

2.1 Polymers of interest

Microplastics used in the deployment experiments were high-density PE (Rigidex

HD6070EA, Goodfellow Cambridge Ltd.); rigid PVC (Geon™ Vinyl Rigid Extrusion

6935, PolyOne Corp.); PLA (100% Natural Virgin, IC3D, LLC); and PHB (Mirel DP9002,

Metabolic Corp.). Chitin (Chi) was harvested from Callinectus sapidus specimens collected from the York River. The carapaces of four previously frozen blue crabs were removed, boiled for 30 minutes and scrubbed with a sterile toothbrush to sanitize the shell and remove any epiphytic coating. Polymer physicochemical characteristics are presented in Table 1, including surface characteristics.

2.2 Sample Deployment & Collection

Two grams of PE, PVC, PHB, PLA beads (2-4 mm in diameter) and Chi fragments were placed in individual fiberglass mesh bags and deployed in the York River Estuary

(37°14'49.1" N, 76°30'03.5” W) from July 11 to August 8, 2017. Previous work in the estuary has indicated the highest concentration of bacterioplankton cells were present during the summer months (Schultz. et al. 2003). Bags were attached to a buoy line at the water surface or lower in the water column within the photic zone in accordance with each polymer’s specific gravities (i.e. PE, PLA, PHB, Chi, PVC). Triplicate samples for each polymer were retrieved during low tide after 7 days, 14 days, and 28 days of immersion. Water quality data were obtained from a continuous water quality monitoring station maintained by VIMS near the deployment site (37°14'50.2'' N, 76° 29' 57.7'' W).

These data are provided in Table 2. The mean water temperature during the full 28-day deployment period was 27.7 ± 0.02°C and average salinity was 20.8 ± 0.02 ppt.

After retrieval, each bag was rinsed with deionized water to remove debris and large colonizing animals, including amphipods, tunicates and juvenile crabs. The bags were cut open and MPs or Chi fragments were collected with sterilized forceps, avoiding tunicates and polychaetes present in the bags. Collected substrate were transferred to 50mL Falcon tubes and stored at -20°C until DNA extractions were performed.

2.3 DNA Extraction & Amplification

Ten MP beads or an equivalent weight of Chi fragments (~1g) were randomly selected from each polymer type and used for DNA extraction with the DNEasy

PowerSoil kit (QIAGEN GmbH, Germany), following the manufacturer’s instructions. To account for the larger size of the beads, MP samples were added to 15 mL Falcon tubes followed by the contents of two PowerSoil garnet bead-beating tubes in order to adequately cover the MPs. Biofilms on the substrates were disrupted by vortexing for 10 minutes. The extracted DNA was measured with Qubit 2.0 and quality was evaluated

with polymerase chain reaction (PCR) of 16S rRNA genes using 27F (5’-

AGAGTTTGATCCTGGCTCAG, Frank et al. 2008) and 685R primers (5’-

ATCTACGGATTTCACTCCTACA, Stults et al. 2001).

2.4 Next-Generation Sequencing of 16S rRNA genes

16S rRNA genes of the biofilm communities were amplified using the forward primer

515F-Y (5’- GTGYCAGCMGCCGCGGTAA, Parada et al., 2016) and reverse primer

926R (5’-CCGYCAATTYNTTTRAGTTT, Quince et al., 2011) encompassing the V4 and V5 hypervariable regions. The PCR reaction contained 10 μL of Go-Taq mix, 1 μL of primers at 10 μM, 1 μL of template DNA (10–30 ng/μL), and nuclease-free H2O up to 25

μL. PCR conditions included a 3 min initial denaturation step at 95°C, followed by 25 cycles of 95°C for 30s, 55°C for 1 min, 72°C for 1 minute, and a final extension of 72°C for 5 minutes. Successful amplification was verified using 1% agarose gel electrophoresis. The gel was stained using a 1% ethidium- bromide solution and imaged under UV light.

PCR products were purified using a MagBio HighPrep PCR clean-up kit. Amplicon concentrations were quantified by Qubit 2.0 using the 1x dsDNA HS Assay kit, and then were diluted to 0.2 ng/mL. The purified amplicons were used for index PCR and then sequenced using the paired-end 500-cyle kit and V3 chemistry according to the Illumina

MiSeq platform protocols for 16S rRNA genes.

2.5 Sequence Processing and Classification

Initial quality filtering and binning were performed to generate fastq files using

Illumina-Utils (Eren et al. 2013), including removal of adapter sequences from the primers. The DADA2 package (v. 1.8.0; Callahan et al. 2016), implemented in R

statistical software (v. 3.5.1; R Core Team 2018) was used to inspect quality of reads, filter and trim reads, dereplicate reads, denoise, merge reads, remove chimeric sequences, and cluster sequences in Amplicon Sequence Variants (ASVs). Taxonomic classification of ASVs was conducted based on the SILVA v. 132 (Quast et al. 2012) reference database for DADA2 at 80% as minimum similarity.

2.6 Data Analysis and Statistics

The phyloseq package (v. 1.24.2; McMurdie & Holmes 2013) in R was used to remove the ASVs of archaea, eukaryotes, chloroplasts, and mitochondria, generating

1754 ASVs in 46 samples. Phyloseq was also used to rarefy the sequences, and examine diversity and composition of the biofilm communities on the different substrates. Low prevalence taxa were removed if they occurred less than once in 2.5% of all samples resulting in 558 ASVs across all samples. Rarefaction was verified by plotting number of

ASVs against the number of reads using the vegan package (v. 2.5-3; Oksanen et al.

2018). Venn diagrams showing shared ASVs across substrates were generated using the package Venn (v. 1.7; Dusa 2018). An NCBI BLAST search (Altschul et al. 1990) was used to identify shared ASVs for potential hydrocarbonoclastic or plastic-degrading species, and pathogenic species. Sampling replicates were pooled and rarefied to the minimum number of sequences for relative abundance bar plots and cluster analysis (n =

15,815). Hierarchical cluster analysis was performed using Jaccard distances and complete-linkage clustering (farthest neighbor clustering) in the base R stats package (v.

3.5.1), and SIMPROF analysis was performed to determine significantly different clusters (clustsig, v.1.1; Whitaker & Christman 2014). For ordination, alpha- and beta- diversity analysis, unpooled replicate samples were rarefied to the smallest number of

sequences in an individual replicate (n = 3,176). Alpha-diversity measures were compared across polymer and treatment types and deployment times using ASVs observed, Chao1, Shannon, and Inverse Simpson diversity indices by one-way ANOVA.

Tukey’s Honestly Significant Difference (HSD) test was used as a post hoc test to confirm ANOVA results. Jaccard distances between unpooled samples were visualized with principal coordinate analysis (PCoA). Statistical comparison between polymer types and dates was done by permutational multivariate analysis of variance (PERMANOVA) with the adonis function in vegan and followed up with permutational pairwise comparisons (pairwiseAdonis v. 0.2; Arbizu 2017). Homogeneity of group dispersions were subsequently tested using betadisper (vegan) to ensure robustness of the

PERMANOVA. A p-value of 0.05 was set as the significance level for all analyses.

3. Results

3.1 Taxonomic trends over time

Figure 1 shows the relative abundance of taxa at the family level. Table 3 depicts the relative abundances of the top ten families overall by substrate. Three families remained prominent throughout the 28-day immersion period; Bacillales Family XII

(Exiguobacterium sp.), Pseudomonadaceae, and Planococcaceae (Table 4). At the first sampling time, Day 7 (Table 5), Rhodobacteraceae was more discriminant of bio-based

MPs (37.4% on PHB-7, 10.2% PLA-7, 8.7% PE-7, 6.2% PVC-7, 2.8% Chi-7), and

Erythrobacteraceae discriminant of MPs overall, in particular petro-based MPs (33.4%

PVC-7, 25.5% PE-7, 18.7% PLA-7, 16.2% PHB-7, 0.7% Chi-7). By Day 14 (Table 6),

Bacillaceae dominated PLA-14 (65.0%) and PE-14 (43.6%), and was present at lower relative abundances across the other MPs only (PVC-14 10.0%, PHB-14 4.5%).

Halomonadaceae was detected at a particularly high abundance (24.8%) on PHB-14, and to a lower extent on PVC-14 (10.4%), and Chi-14 (3.6%). Colonization by

Paenibacillaceae was also noticeable (14.7% PVC-14, 7.5% Chi-14, 5.1% PLA-14, 4.5%

PE-14, 3.2% PHB-14). Interestingly, Erythrobacteraceae became dramatically less abundant overall (6.6% PHB-14, 4.4% PE-14, 1.9% Chi-14, 0.2% PLA-14, 0.1% PVC-

14). In addition, Vibrionaceae was detected at a prominent level only on Chi on Day 7

(4.6%) and Day 14 (4.1%). As the biofilms began to stabilize by Day 28 (Table 7),

Xanthomonadaceae dominated PVC -28 (37.0%) and colonized Chi-28 (1.1%), PE-28

(0.4%), and PHB-28 (0.3%). Erythrobacteraceae became prominent on Chi-28 (18.2%),

PVC-28 (10.5%), and PLA-28 (5.1%); remaining steady on PE-28 (5.3%), and decreasing on PHB-28 (1.4%).

3.2 Generalist vs. specialist on different MP polymers

To examine the communities between the common natural polymer, chitin, and the bio-based and petroleum-based synthetic polymers, substrates were grouped into treatments (i.e. Control, Bio, and Petro). The number of shared and individual ASVs between the natural polymer Chi and Bio- and Petro-based MPs is shown in Figure 2, and top ten genera are listed in Table 8. Abundances listed encompass all sampling times and substrates. Out of 558 ASVs, 60 were generalists shared between all substrate types. The most abundant ASV was identified as Exiguobacterium aestuarii. (16.0%), followed by

Paracoccus sediminius (6.3%), and Pseudomonas zhaodongensis (5.4%).

Exiguobacterium sp. and P. sediminius were ubiquitous across all polymers and all sampling times (Day 7, Day 14, Day 28). P. zhaodongensis was primarily identified on

Day 28 biofilm samples. Rare taxa defined as <0.01% of relative abundance of ASVs found on all substrates include Nioella aestuarii and Oceanicola antarcticus.

The common generalists on synthetic plastic substrates (i.e. PE, PVC, PLA, PHA) consisted of the 25 ASVs shown in Figure 2 and Table 8. The most abundant of these were Paenibacillus populi (2.55%), Bacillus australmaris (0.54%), and Bacillus subterraneus (0.13%). B. subterraneus was identified only during Day 7 sampling, and B. australmaris only during Day 14. P. populi was identified during each sampling point but was the most prominent on Day 14. The rare members included a member of the family

Erythrobacteraceae, Paenibacillus turicensis, and Ilumatobacter nonamiensis.

Fourteen ASVs were specialists shared exclusively by the petro-based synthetic polymers (Figure 2, Table 8). The most abundant taxa included Rheinheimer aquimaris

(0.24%), Noviherbaspirillum suwonense (0.24%), Bacillus isronensis (0.08%), and

Pseudoxanthomonas sp. (0.07%). Each of these was identified as an exclusive member of the mature biofilm community (Day 28 samples only), and R. aquimaris was identified mainly on PVC (0.239% v. 0.002% on PE). Half of the shared ASVs could be considered rare microbes (<0.01%). These included early colonizers only found during the first week: Paenibacillus vini, Erythrobacter aquimaris, Paenibacillus ginsengarvi, and a member of the family Phyllobacteriaceae.

Of the 32 specialist ASVs only shared between only PHB and PLA (Figure 2, Table

8), the most abundant were Sporosarcina sp. (0.35%), Psychrobacillus soli (0.26%), and

Streptomyces sanyensis (0.23%). S. sanyensis was identified as a specialist on PHB from

Day 7 (0.22% v 0.01% on PLA). Sporosarcina sp. was only found during Day 14

samples. Twenty of the 32 ASVs are rare taxa, including Erythrobacter vulgaris,

Salinova rubellum, and Castellaniella sp..

Thirty-nine ASVs were found to be specialists on Chi (Figure 2, Table 8). The most abundant were: Timonella senegalensis (0.73%), Fusibacter tunisiensis (0.56%), and

Vibrio sp. ASV 125 (0.41%). F. tunisiensis was identified exclusively on Day 7 samples,

T. senegalensis on Day 14, and Vibrio sp. ASV 125 was present at both sampling points, but in greater abundance on Day 7. Rare taxa included Clostridium carnis, Clostridium paraputrificum, and Paenibacillus sp. Both Clostridium spp. were identified during Day

14, and Paenibacillus sp. during Day 28 sampling.

3.3 Cluster Analysis

Clustering analysis of pooled replicates (Figure 1) showed distinct dissimilarity between Chi samples (distance = 0.663), Day 7 MPs (0.709), Day 14 MPs (0.810), and

Day 28 MPs (0.828), with a few exceptions. PHB-14 and Chi-28 both clustered with the

Day 28 samples. For the Day 7 cluster of samples, PE-7 clustered more closely with

PHB-7 and PVC-7 clustered more closely with PLA-7. The same holds true of Day 28, where the PVC-28 and PLA-28 pair show the highest dissimilarity overall (0.601). Day

14 samples form a separate clade from the other groups.

3.4 Alpha diversity

Changes in alpha diversity were related to sampling time, with higher overall indices for Day 28 samples (Figure 3, Table 9). PHB showed higher richness than the other polymers over Day 7 and Day 14, with PVC having the highest richness on Day 28. One- way ANOVA was used to determine if differences in indices were statistically significant

(Table 10), with the overall subset indicating differences by Day. Day 7 demonstrated no

statistical differences across any of the diversity measures tested between substrates, but significant differences were detected at the treatment level for Observed ASVs (p =

0.0172) and Shannon index (p = 0.0298). Day 14 showed a statistical difference in

Shannon (p = 0.0161) and Inverse Simpson (p = 0.0090) indices between substrates, and a statistical difference at the treatment level for Inverse Simpson. One-way ANOVA on

Day 28 substrates and treatments showed no statistical difference in any alpha diversity index. Tukey’s HSD post-hoc test was performed on those factors that were significant

(Table 11).

3.5 Beta diversity

Principal coordinate analysis (PCoA) of Jaccard distances (Figure 4) followed the cluster analysis closely, with samples grouping together by date and substrate type.

Overall, Day 14 showed the greatest dissimilarity along PCo 1 (10.3%), and Day 7 along

PCo 2 (8.8%). PHB-14 clearly clustered with the Day 28 polymers, and Chi-7 and Chi-14 form their own cluster in the lower left-hand corner of the plot.

PERMANOVA (Table 12) was performed only on Day 7 and Day 28 samples, due to the loss of two replicates during Day 14 and the desire to focus in on differences between early (Day 7) and established (Day 28) biofilm communities. Results confirmed the

PCoA and cluster analysis, with significant differences found between Substrate and Date

(p = 0.001; Table 12, ‘All’). A significant interaction between Substrate and Date was also detected by PERMANOVA. Investigation of the betadisper plot (Figure 5), showed that this interaction was likely caused by Chi-7 clustering far from the other Day 7 samples and interacting Day 28. At the treatment level (Table 12, ‘Treatment’), there were significant differences in microbial communities between each pair of treatments,

and Date was also a significant factor. Pairwise comparison (Table 13) of the interaction between Date and Treatment showed only significant interaction for the chitin-including pairs of treatments, confirming what was inferred from betadisper plot. The communities on MPs varied significantly by polymer type and by date (Table 12), with greater variation between the two petro-based polymers (PE – PVC; p = 0.002) than the bio- based polymers (PLA – PHB; p = 0.05). For the petro-based plastics, the interaction between date and polymer was also significant (p = 0.005). Permutational dispersion test was used to verify homogeneity of variance across all factors using betadisper, finding three instances where variance was heterogenous and the assumptions of PERMANOVA violated (Table 14a, bold).

Sub-setting PERMANOVA analysis by day provided a clearer picture (Table 12b).

Looking just at Day 7, there were significant difference between all substrates (p =

0.001), and each pair-wise comparison of treatments, but no statistical difference between polymers within treatments (i.e. Bio: PLA – PHB; Petro: PE – PVC). By Day 28, there continued to be a statistical difference between all substrates (p = 0.001), and between

Control – Bio (p = 0.011) and Control – Petro (p = 0.023) treatments levels, but no difference between Bio and Petro polymers (p = 0.169). There was no statistical difference between polymers within treatment groups at Day 28. Homogeneity of variance across all factors was verified using betadisper (Table 14b).

4. Discussion

4.1 Bacterial Biofilm Dynamics – early colonization to pseudo-stabilization

Progressing from early colonization (Day 7) through the growth phase (Day 14) to stabilization (Day 28), the Jaccard index of each Date cluster increased from 0.709 to

0.810 and finally to 0.828. This indicated that the communities on each substrate within each date-separated clade became more similar over time. This suggests that early differences in biofilm communities resolve. This is further supported by the spread of samples on the PCoA (Figure 4), with the most variation seen in Day 7 and Day 14, and closer clustering overall on Day 28. This was expected, as the early colonization and growth phase of a biofilm is characterized by a high level of flux between community members (Fisher et al. 2014). Chitin appears to be driving the differences in the biofilm communities among substrates. I examined the differences between chitin and the other substrates more in the following sections. In terms of diversity, Chi follows a clear trend of increasing diversity measures as time progresses, whereas the MPs vary in their responses. Diversity drivers in marine biofilms revolve around environmental factors such as nutrient availability, light, temperature, pH, CO2, dissolved O2, and influx of new species (through tidal movement, runoff, estuarine mixing), as well as quorum sensing among colonists (Jain & Bholse 2010). It was beyond the scope of this project to further investigate environmental parameters and changes in diversity measures.

Taken together, the incubation time was a greater driver of biofilm community than substrate type. Other studies that investigated microbial communities across different substrates have found that spatio-temporal factors to be greater than substrate differences in shaping biofilms over 2 – 6 weeks (Oberbeckmann et al. 2016; Oberbeckmann et al.

2018; Ogonowski et al. 2018), with no significant differences to be found in communities between substrate types.

Of the studies that compared early biofilms to later stage biofilms, De Tender et al.

(2017) investigated two types of PE (sheeting and rope) over a 44-week period and found

that communities gradually evolved over the 44 weeks. Only minor differences were noted between the two plastic types across that time frame. Alternatively, Dussud et al.

(2018) examined early stage through stabilization of biofilm formation, finding community similarities at day 7 and differences between some polymer substrates at later sampling dates (22, 30 and 45 days). Interestingly, in the Dussud study, LDPE and OXO-

PE showed distinct communities from one another, even though the material type were generally the same. The aged-OXO and PHVB communities converged to a large extent by day 45, aligning with the findings of other studies (Oberbeckmann et al. 2018;

Ogonowski et al. 2018; Dussud et al. 2018b).

De Tender et al. and others note that Alpha- and Gammaproteobacteria are characteristic for primary biofilm colonization on submerged surfaces, with Bacteroidetes heralding secondary colonization (Lee et al. 2008; Elifantz et al. 2013). In the current study, Alphaproteobacteria (Rhodobacteraceae, Sphingomonaceae/Erythromonadaceae),

Gammaproteobacteria (Pseudomonadaceae) and Bacilli (Bacillaceae, Planococcaceae,

Bacillales Family XII [Exiguobacterium sp.]), were the dominating phyla throughout the

4-week incubation period. Abundances of Bacteroidetes (Flavobacteriaceae) remains relatively low throughout the timeframe investigated, peaking on Chi on Day 28 (5.5%).

In a recirculating natural seawater system, Dussud et al. found Gammaproteobacteria were the dominant early colonizers (day 7), but these decreased in abundance over time.

This was accompanied by a proportional increase in Alphaproteobacteria during growth

(> 7 days) and maturation of the biofilm (> 22 days). Chi from this study followed a similar trend, with decreased abundance of Gammaproteobacteria and increased

Alphaproteobacteria on Day 28.

4.2 Comparison of biofilm communities on natural and synthetic polymers

PCoA showed a clear delineation between chitin and the synthetic polymer communities during Day 7 and Day 14, but the chitin group began to approach the others by Day 28 (Figure 4). This follows what was surmised investigating the dynamics of biofilm communities over time; biofilm communities become more similar over time, despite early differences. The differences between Chi and the other treatments appeared to be driven by differences in Proteobacteria, particularly reduced Bacillaceae,

Erythrobacteraceae, Pseudomonadaceae, and higher abundances of Planococcaceae.

Chitin also was the primary harborer of Vibrionaceae (0.3% Chi vs. 0.005% PVC). We discuss the presence of pathogenic bacteria further on.

Those that have considered naturally occurring organic material have reported distinct differences between these materials (i.e. leaves, organic particles), and incubated plastics

(Hoellein et al. 2014) or MPs collected from the environment (McCormick et al. 2014).

In these studies, natural substrates were distinguished from plastic by an increased abundance of Rhodocyclaceae and Thiotrichaceae on particle-attached communities

(McCormick et al. 2014), or by Oxalobacteraceae, Enterobacteraceae, and Rhizobiaceae on leaves (Hoellein et al. 2014). These two studies were carried out in freshwater environments. Oberbeckmann et al. (2018) investigated the biofilm communities associated with PE, polystyrene (PS), wood and naturally present particles in the Baltic

Sea, finding a significant difference between individual plastic and wood or particle- attached communities. The authors did not comment on which taxa were more abundant on wood or PA communities and focused instead on the plastic associated biofilms. In this regard, they found the families Hyphomonadaceae, Plantomycetaceae,

Erythrobacteraceae, and Sphingomonodaceae to be discriminant of plastic communities

(Oberbeckmann et al. 2018). In this study, Chi also showed higher abundances of

Rhizobiaceae and polymer communities were selective for Erythrobacteraceae (with the exception of Chitin from Day 28). Datta et al. (2016) investigated the micro-scale successions of microbes on the surface on paramagnetized chitin particles, finding colonization by the orders Vibrionales, Alteromonadales, Oceanospirallales,

Rhodobacterales, and Flavobacteriales present at > 1% after six days. I found each of these taxa present at Day 7 with Vibrionales (Vibrionaceae), Alteromonadales

(Marinobacteraceae), and Rhodobacterales (Rhodobacteraceae) present at > 1% (Table

3). These orders were also identified at the other time points. Ogonowski et al. (2018) investigated communities on cellulose and glass as compared to PE, PS, and PP. They found that plastic and non-plastic substrates formed distinct groupings in PCoA, though statistical analysis was not performed. Unclassified Actinobacteria genera were discriminative of cellulose and glass in this study, but this phylum was common among all substrates in this study. On chitin, the most abundant genera of Actinobacteria were

Timonella sp., Streptomyces sp., and Rhodococcus sp.

4.3 Comparison of biofilm communities between bio-based and petrochemical-based polymers

Initial communities on bio-based and petro-based polymers showed a statistical difference which converged by the Day 28 sampling (Table 12). PERMANOVA also verified the classification of PE and PVC as a specific “treatment” of polymer (Petro), with PHB and PLA as another (Bio), though cluster analysis did not group the two treatments of polymers into separate clades (Figure 1). Bio-based MP communities had a

higher abundance of Rhodobacteraceae, Bacillaceae, and Planococcaceae. Petro MPs had a higher abundance of Xanthomonadaceae, Sphingomonadaceae and

Caulobacteraceae. Generalists for both groups include Paenibacilius populi.

Dussud and colleagues (2018) investigated the microbial community differences between degradable and non-biodegradable plastics. They found that PHBV, a degradable, bio-based polymer, was readily colonized most heavily by

Oceanospirallaceae (Neptuniibacter sp.) and by Rhodobacteraceae (Phaeobacter sp.). In this study, Rhodobacteraceae accounting for 37.4% and 10.2% of total abundance on

PHA and PLA on Day 7. Predominant families of LDPE and OXO-LDPE included

Alcanivoraceae, Alteromonadaceae, Oceanospirallaceae, Cellvibrionaceae,

Flavobacteriaceae, Sneathiellaceae, and Sinobacteraceae. Aged OXO-LDPE was predominantly colonized by Sphingomonadaceae, Oceanospirallaceae, Alcanivoraceae, and Alteromonadaceae. In this study, PE and PVC had some presence of

Sphingomonadaceae, but had a high abundance of the closely related family

Erythrobacteraceae (25.5% on PE; 33.4% on PVC – Day 7). Low abundances of

Flavobacteraceae and Alteromonadaceae were also detected. None of the other families were detected on our Petro-based samples. Interestingly, PHB and PLA showed significant abundance of Erythrobacteraceae (16.2% and 18.7% on Day 7, respectively), indicating that Erythrobacteraceae may be able to universally colonize and perhaps utilize such polymers as a carbon source. Tonon et al. (2014) noted that this family contains several putative hydrocarbonoclastic bacteria. Erythrobacteraceae were also reported to be abundant on PE samples deployed in estuarine conditions in the Baltic Sea

(Oberbeckmann et al. 2018).

4.4 Presence of hydrocarbonoclastic and plastic degrading bacteria

Several studies on plastic-associated bacterial communities have identified hydrocarbonoclastic bacteria (HCB) and plastic-degrading bacteria (Zettler et al. 2013;

McCormick et al. 2014; Oberbeckmann et al. 2016; De Tender et al. 2017; Dussud et al.

2018a, 2018b; Oberbeckmann et al. 2018). These bacteria are capable of degrading complex carbon substrates due to the presence of specialized genes for alkane degradation (Brooijmans et al. 2009), and these capabilities may be applicable to plastic- degradation. These genes include hydroxylases, mono- and dioxygenases, and cytochrome P450 enzymes (Abbasian et al. 2015, Brooijmans et al. 2009). HCBs also tend to produce exopolysaccharides and biosurfactants that help attach them to hydrophobic surfaces and emulsify oils (Brooijmans et al. 2009). In this study, evidence of several HCB genera was found (Table 15). Specialists of petro-based MPs included suspected hydrocarbonoclastic bacteria Rheinheimera aquimaris, Pseudoxanthomonas mexicana, and Erythrobacter aquimaris.

The most abundant genus across all substrates was Exiguobacterium. Three ASVs of

Exiguobacterium were identified: ASV 1, ASV 9, and ASV 40, with ASV 1 and ASV 9 found on all substrates and at all sampling times. Of the BLAST species matches (Table

6, Table 11), some have been identified as key in bioremediation of hydrocarbons

(Kasana & Pandey 2017), and others have been identified as degraders of hydrocarbons

(E. auranticum: Mohanty & Mukherji 2005; Exiguobacterium sp.: Kumar et al. 2014) and degraders of chitin (E. acetylicum: Kumar & Suresh 2014). Two species of

Exiguobacterium have been identified as using PS as a carbon source (Chauhan et al.

2018). Additionally, a strain of Exiguobacterium related to E. indicum has been isolated

from the midgut of a mealworms fed PS (Yang et al. 2014). Another study isolated an

Exiguobacterium sp. strain capable of degrading a starch/HDPE blend (Muthukumar et al. 2014). This demonstrates that Exiguobacterium spp. may degrade polymers, yet no other study investigating MP-associated microbes have explicitly identified this genus.

Exiguobacterium is a member of Firmicutes, which are typically associated with wastewater and sewage (Oberbeckmann et al. 2015), so studies utilizing plastics deployed in or collected from the open ocean (e.g. Zettler et al. 2013; Oberbeckmann et al. 2014) are unlikely to find evidence of this genus. One open ocean study (Carson et al.

2013) did identify “Bacillus” type bacteria from scanning electron micrographs of MPs collected from the North Pacific Gyre, but molecular analysis was not performed to verify the authors identification.

4.5 Pathogenic bacteria

Some early studies of microbial populations on PMD indicated an increase in the abundance of pathogenic bacteria (Zettler et al. 2013; McCormick et al. 2014;

Oberbeckmann et al. 2015). Potentially pathogenic bacteria identified in this work are listed in Table 16.

Of particular concern with respect to pathogenic bacteria is Vibrio spp., as Zettler and coauthors identified Vibrio sp. on a PP sample from the North Atlantic at nearly 24% abundance (2014). Frère et al. also found an abundance of Vibrio sp. (9.9%) on microplastic samples in the Bay of Brest (2018). In the present study, three ASVs of

Vibrio spp. were identified; ASV 50, ASV 125 and ASV 166. Each ASV was found on

Chi during the first and second week of immersion, and ASV 50 was also identified on

PVC during Day 7 sampling. BLAST search returned several 100% matches for each

ASV, and hence further identification could not be verified (Table 12). This is a known issue with Vibrio sp. identification by 16S rRNA (Reen et al. 2006; Machado & Gram

2015). Included in these matches were several known human pathogenic Vibrio spp. including V. parahaemolyticus and V. alginolyticus, and the fish pathogen V. harveyi. In total, these ASVs accounted for 0.27% of total abundances across sampling times and substrates. Individually, ASV 50 was present as high as 3.8% on our Chi sampled at Day

7, and 2.2% at Day 14. ASV 166 was also abundant at Day 14 sampling (1.4%). This is significant as Vibrio sp. are typically present at < 1% (Zettler et al. 2013) in the water column. There were no increased Vibrio sp. abundances on plastic substrates, which has been the case in a few studies (Debroas et al. 2017; De Tender et al. 2017;

Oberbeckmann et al. 2018). Instead Vibrio spp. were harbored on the naturally occurring chitin. Oberbeckmann et al. (2018) found increased Vibrio spp. abundance on wood particles and naturally occurring particles over plastics. Debroas (2017) hypothesized that

Vibrio are not involved in plastic degradation but are opportunistic hitchhikers on PMD, and my results tend to agree with this assessment.

Five ASVs of the pathogenic Mycobacterium were also recognized in the present study. Some species of Mycobacterium have been associated with human mycobacteriosis skin lesions, and lesions in fish, including striped bass in the Chesapeake

Bay. Interestingly, Mycobacterium ASVs were more associated with MPs than chitin, particularly PLA (2.5% abundance at Day 7). Mycobacterium have also been identified in other works (Dang & Lovell 2000; McCormick et al. 2014).

5. Conclusions

This research examined the microbial biofilm communities of five polymer types over time, finding that any early differences in biofilm communities converged over the course of the incubation. These early differences were driven primarily by significant differences in the biofilm colonizing chitin, a naturally occuring, abundant biopolymer.

For synthetic polymers, early differences observed between conventional, petrochemical- based MPs and bio-based MPs also resolved over 28 days. Based on this work, it seems that PHB reached a stable biofilm community more quickly than the other polymers

(Figure 3), as PHB reached maximum richness by the second week of sampling (Day 14) and dropped off precipitously at Day 28. This drop could be the result of biofilm dispersal that accompanies maturation (Dobretsov 2010). This seemingly faster maturation could be due to the hydrophilic nature of the PHB surface allowing for rapid colonization, or perhaps an artifact of the wide-spread capability for bacteria to degrade

PHB leading to rapid turnover.

With regards to hydrocarbon and plastic-degrading bacteria, I identified

Exiguobacterium spp. as a primary colonizer in the estuarine environment studied.

Pathogenic bacteria such as Vibrio spp. were not found in exaggerated abundance on

MPs, though Mycobacterium spp. preferentially colonized PLA.

Based on evidence from other works, bacterial communities seemed more determined by the surrounding environment than on substrate type. This calls into question that a novel “Plastisphere” core microbiome exists. However, Hyphomonaceae,

Erythrobacteraceae and Tenacibaculum sp. have been documented on MPs in several studies spanning the North Atlantic, North Pacific, North Sea, and the Baltic Sea (Zettler

et al. 2013, Bryant et al. 2016; Dussud et al. 2018; Oberbeckmann et al. 2018). I also observed occurrences of these families (particularly Erythrobacteraceae), perhaps speaking more to the ubiquity of these bacteria than the deterministic power of microplastic debris.

6. Tables

Table 1. Polymeric substrates used in the 28-day estuarine immersion experiment.

contact density angle w/ ID crystallinity shape structure (g/cm3)a water (°)

elliptical HDPE 0.96 135.6c 58.5%c cylinder

83.2 - rectangular PVC 1.53 4 – 10 %a 91.9a prism

elliptical PHB 1.22 75d 60 %f cylinder

PLA 1.24 77e 37%f ellipsoid

Chi 1.14b – – irregular

aWypych 2012; bBlake 1985; cGuo et al. 2012; dKang et al. 2001; eParagkumar et al. 2006; fRoohi et al. 2018

Table 2. Water quality for the length of the immersion experiment from a nearby continuous monitoring station.

July 11 - July 18 July 18 - July 25 July 11- July 25 July 25 - August 8 July 11- August 8 Day 7 Day 14 Day 28 parameter Cumulative Day 7 - 14 Cumulative Day 14 - 28 Cumulative Temp (°C) 27.81 ± 0.028 28.71 ± 0.027 28.28 ± 0.023 27.19 ± 0.019 27.73 ± 0.018 Salinity (ppt) 20.90 ± 0.042 20.35 ± 0.042 20.62 ± 0.031 21.09 ± 0.033 20.85 ± 0.023 pH 7.77 ± 0.006 7.87 ± 0.005 7.82 ± 0.004 7.99 ± 0.006 7.91 ± 0.004 DO (mg/L) 5.74 ± 0.051 6.32 ± 0.040 6.04 ± 0.033 7.55 ± 0.059 6.79 ± 0.037 DO SAT (%) 82.17 ± 0.755 91.61 ± 0.603 87.06 ± 0.497 107.14 ± 0.854 97.07 ± 0.530 TURB 7.14 ± 0.108 8.33 ± 0.199 7.76 ± 0.117 9.28 ± 0.592 8.52 ± 0.301 Fluor (% FS) 1.83 ± 0.018 2.14 ± 0.028 1.99 ± 0.017 4.20 ± 0.195 3.09 ± 0.100 Total ChlA (ug/L) 6.83 ± 0.066 7.53 ± 0.096 7.19 ± 0.060 15.37 ± 0.725 11.27 ± 0.371

Table 3. Dominant 10 families and percent abundances by substrate over all days sampled.

Substrate Family % Substrate Family % PE Bacillales Family XII 33.4% PVC Pseudomonadaceae 20.2% Bacillaceae 17.0% Xanthomonadaceae 19.0% Erythrobacteraceae 14.3% Erythrobacteraceae 11.2% Rhodobacteraceae 9.3% Bacillales Family XII 11.2% Pseudomonadaceae 6.6% Caulobacteraceae 8.1% Planococcaceae 5.9% Planococcaceae 6.7% Paenibacillaceae 3.0% Paenibacillaceae 4.9% Caulobacteraceae 3.0% Rhodobacteraceae 4.5% Streptomycetaceae 1.2% Bacillaceae 3.7% Others 6.3% Others 10.4% PHB Rhodobacteraceae 17.6% PLA Bacillales Family XII 21.9% Bacillales Family XII 15.8% Bacillaceae 21.7% Halomonadaceae 11.2% Pseudomonadaceae 11.8% Pseudomonadaceae 23.9% Planococcaceae 10.8% Planococcaceae 7.9% Erythrobacteraceae 9.0% Erythrobacteraceae 7.7% Rhodobacteraceae 5.9% Burkholderiaceae 2.6% Caulobacteraceae 3.7% Microbacteriaceae 2.3% Paenibacillaceae 2.9% Paenibacillaceae 1.9% Clostridaceae I 1.2% Streptomycetaceae 1.8% Mycobacteraceae 1.2% Others 7.2% Others 9.7% Chitin Bacillales Family XII 40.1% Overall Bacillales Family XII 19.5% Planococcaceae 20.4% Pseudomonadaceae 16.6% Erythrobacteraceae 8.3% Erythrobacteraceae 10.0% Pseudomonadaceae 5.6% Planococcaceae 9.1% Rhodobacteraceae 4.6% Rhodobacteraceae 8.3% Flavobacteriaceae 3.0% Xanthomonadaceae 7.5% Vibrionaceae 2.5% Caulobacteraceae 4.1% Rhizobiaceae 1.7% Halomonadaceae 4.0% Marinobacteraceae 1.6% Paenibacillaceae 3.3% Paenibacillaceae 1.4% Flavobacteriaceae 1.4% Others 10.7% Others 16.3%

Table 4. Dominant 10 Families and percent abundances on all substrates over all days sampled.

Sampled Genus Abundance Day 7 Bacillales Family XII 31.5% Erythrobacteraceae 19.9% Rhodobacteraceae 13.8% Pseudomonadaceae 11.2% Planococcaceae 9.6% Burkholderiaceae 2.1% Bacillaceae 2.0% Xanthomonadaceae 1.3% Caulobacteraceae 1.2% Rhizobiaceae 0.9% Others 6.4% Day 14 Pseudomonadaceae 23.0% Bacillaceae 20.5% Bacillales Family XII 13.8% Halomonadaceae 11.9% Paenibacillaceae 8.0% Rhodobacteraceae 7.3% Planococcaceae 6.2% Erythrobacteraceae 2.7% Microbacteriaceae 1.8% Streptomycetaceae 1.1% Others 3.8% Day 28 Xanthomonadaceae 18.9% Pseudomonadaceae 15.2% Bacillales Family XII 15.1% Planococcaceae 11.2% Caulobacteraceae 10.0% Erythrobacteraceae 8.7% Rhodobacteraceae 5.0% Flavobacteriaceae 2.6% Bacillaceae 1.8% Paenibacillaceae 1.4% Others 10.1%

Table 5. Dominant 10 Families and percent abundance on all substrates sampled during

Day 7.

PE Abundance PVC Abundance Bacillales Family XII 42.1% Erythrobacteraceae 33.4% Erythrobacteraceae 25.5% Bacillales Family XII 19.9% Rhodobacteraceae 8.7% Pseudomonadaceae 19.4% Bacillaceae 5.5% Planococcaceae 14.0% Pseudomonadaceae 5.4% Rhodobacteraceae 6.2% Caulobacteraceae 3.8% Xanthomonadaceae 2.4% Planococcaceae 3.3% Rhizobiaceae 1.1% Saprospiraceae 1.2% Nocardiaceae 0.8% Paenibacillaceae 0.5% Bacillaceae 0.4% Others 3.9% Others 2.6% PHB PLA Rhodobacteraceae 37.4% Bacillales Family XII 27.5% Bacillales Family XII 23.0% Erythrobacteraceae 18.7% Erythrobacteraceae 16.2% Pseudomonadaceae 15.9% Burkholderiaceae 9.1% Planococcaceae 11.2% Pseudomonadaceae 3.8% Rhodobacteraceae 10.2% Planococcaceae 2.4% Xanthomonadaceae 2.8% Streptomycetaceae 2.4% Mycobacteriaceae 2.7% Bacillaceae 2.1% Caulobacteraceae 2.3% Devosiaceae 1.0% Bacillaceae 2.3% Paenibacillaceae 0.7% Flavobacteriaceae 0.9% Others 1.8% Others 5.4% Chitin Overall Bacillales Family XII 57.3% Bacillales Family XII 31.5% Planococcaceae 17.3% Erythrobacteraceae 19.9% Pseudomonadaceae 7.5% Rhodobacteraceae 13.8% Vibrionaceae 4.6% Pseudomonadaceae 11.2% Rhodobacteraceae 2.8% Planococcaceae 9.6% Rhizobiaceae 2.5% Burkholderiaceae 2.1% Marinobacteraceae 1.8% Bacillaceae 2.0% Ruminococcaceae 1.2% Xanthomonadaceae 1.3% Devosiaceae 1.0% Caulobacteraceae 1.2% Cyclobacteriaceae 0.8% Rhizobiaceae 0.9% Others 3.2% Others 6.4%

Table 6. Dominant 10 Families and percent abundances on all substrates sampled during

Day 14.

PE Abundance PVC Abundance Bacillaceae 43.6% Pseudomonadaceae 45.4% Bacillales Family XII 20.7% Paenibacillaceae 14.7% Rhodobacteraceae 14.8% Bacillales Family XII 14.0% Paenibacillaceae 4.5% Halomonadaceae 10.4% Erythrobacteraceae 4.4% Bacillaceae 10.0% Planococcaceae 4.2% Rhodobacteraceae 3.8% Streptomycetaceae 3.5% Streptomycetaceae 0.9% Rhizobiales Incertae Sedis 1.0% Planococcaceae 0.4% Flavobacteriaceae 0.6% Erythrobacteraceae 0.1% Others 2.7% Others 0.3% PHB PLA Halomonadaceae 24.8% Bacillaceae 65.0% Pseudomonadaceae 18.7% Bacillales Family XII 15.9% Rhodobacteraceae 13.1% Planococcaceae 6.8% Planococcaceae 11.2% Paenibacillaceae 5.1% Bacillales Family XII 7.8% Streptomycetaceae 1.5% Erythrobacteraceae 6.6% Microbacteriaceae 1.3% Microbacteriaceae 4.8% Nocardiaceae 0.9% Bacillaceae 4.5% Promicromonosporaceae 0.8% Paenibacillaceae 3.2% Rhodobacteraceae 0.5% Flavobacteriaceae 1.4% Rhizobiales Incertae Sedis 0.3% Others 3.9% Others 1.8% Chitin Overall Bacillales Family XII 29.5% Pseudomonadaceae 23.0% Planococcaceae 19.0% Bacillaceae 20.5% Pseudomonadaceae 10.4% Bacillales Family XII 13.8% Paenibacillaceae 7.5% Halomonadaceae 11.9% Rhodobacteraceae 4.7% Paenibacillaceae 8.0% Micrococcales 4.4% Rhodobacteraceae 7.3% Marinobacteraceae 4.2% Planococcaceae 6.2% Vibrionaceae 4.1% Erythrobacteraceae 2.7% Halomonadaceae 3.6% Microbacteriaceae 1.8% Flavobacteriaceae 3.0% Streptomycetaceae 1.1% Others 9.6% Others 3.8%

Table 7. Dominant 10 Families and percent abundances on all substrates sampled during

Day 28.

PE Abundance PVC Abundance Bacillales Family XII 32.9% Xanthomonadaceae 37.0% Pseudomonadaceae 17.8% Caulobacteraceae 16.2% Planococcaceae 13.5% Erythrobacteraceae 10.5% Paenibacillaceae 6.1% Planococcaceae 8.2% Caulobacteraceae 5.5% Bacillales Family XII 6.3% Erythrobacteraceae 5.3% Pseudomonadaceae 4.4% Bacillaceae 4.1% Rhodobacteraceae 4.3% Rhodobacteraceae 2.9% Flavobacteriaceae 3.1% Microbacteriaceae 2.4% Micrococcaceae 1.1% Others 9.5% Others 8.9% PHB PLA Pseudomonadaceae 51.4% Bacillales Family XII 20.7% Bacillales Family XII 21.7% Pseudomonadaceae 18.9% Planococcaceae 7.8% Planococcaceae 14.6% Rhodobacteraceae 6.0% Caulobacteraceae 9.7% Streptomycetaceae 3.1% Rhodobacteraceae 5.8% Bacillaceae 2.4% Erythrobacteraceae 5.1% Erythrobacteraceae 1.4% Paenibacillaceae 4.6% Paenibacillaceae 1.1% Clostridaceae I 4.3% Flavobacteriaceae 1.0% Burkholderiaceae 2.3% Burkholderiaceae 0.7% Bacillaceae 1.6% Others 3.5% Others 12.4% Chitin Overall Bacillales Family XII 27.3% Xanthomonadaceae 18.9% Planococcaceae 23.9% Pseudomonadaceae 15.2% Erythrobacteraceae 18.2% Bacillales Family XII 15.1% Rhodobacteraceae 6.2% Planococcaceae 11.2% Flavobacteriaceae 5.5% Caulobacteraceae 10.0% Bacillaceae 2.4% Erythrobacteraceae 8.7% Pseudomonadaceae 1.9% Rhodobacteraceae 5.0% Clostridaceae I 1.5% Flavobacteriaceae 2.6% Microbacteriaceae 1.4% Bacillaceae 1.8% Rhizobiaceae 1.4% Paenibacillaceae 1.4% Others 10.3% Others 10.1%

Table 8. Dominant 10 genera and percent relative abundances on shared substrates over all days sampled.

Substrate ASV # BLAST match Abundance All 1 Exiguobacterium aestuarii 100% 15.99% 4 Paracoccus sediminis 100% 6.33% 5 Pseudomonas zhaodongensis 100% 5.41% 6 Planococcus citreus 100% 5.41% 7 Erythrobacter citreus 100% 4.71% 9 Exiguobacterium artemiae 100% 3.17% 12 Bacillus sp. 100% 2.53% 15 Bacillus drentensis 100% 1.70% 16 Bacillus hwajinpoensis 100% 1.60% 17 Erythrobacter flavus 100% 1.44% Plastics Only 13 Paenibacillus populi 100% 2.55% 28 Bacillus australimaris 100% 0.54% 70 Bacillus subterraneus 99% 0.13% 72 Skermanella aerolata 99% 0.13% 122 Mycobacterium angelicum 99% 0.05% 128 Anderseniella baltica 98% 0.05% 131 Bythopirellula goksoyri 89% 0.05% 152 Synechococcus rubescens 97% 0.03% 155 Sphingopyxis flavimaris 100% 0.03% 171 Altererythrobacter mangrovi 99% 0.03% Petro 71 Rheinheimera aquimaris 100% 0.24% 73 Noviherbaspirillum suwonense 100% 0.24% 126 Bacillus isronensis 100% 0.08% 142 Pseudoxanthomonas mexicana 99% 0.07% 189 Rhodococcus jialingiae 100% 0.04% 236 Roseomonas ludipueritiae 100% 0.03% 245 Sphingomonas hunanensis 99% 0.03% 337 Paenibacillus vini 100% 0.01% 395 Erythrobacter aquimaris 99% 0.01% 438 Paenibacillus ginsengarvi 100% 0.01%

Table 8 cont.

Substrate ASV # BLAST match Abundance Bio 61 Sporosarcina luteola 100% 0.35% 77 Psychrobacillus soli 100% 0.26% 84 Streptomyces sanyensis 100% 0.23% 102 Devosia subaequoris 99% 0.12% 109 Gracilibacter thermotolerans 95% 0.13% 121 Cellulosimicrobium marinum 100% 0.11% 139 Paenibacillus glucanolyticus 99% 0.09% 157 Microbacterium esteraromaticum 100% 0.06% 159 Bacillus panacisoli 100% 0.05% 195 Brevibacillus fluminis 100% 0.04% Chitin 94 Timonella senegalensis 99% 0.73% 106 Fusibacter tunisiensis 99% 0.56% 125 Vibrio sp. 100% 0.41% 130 Halomonas nanhaiensis 99% 0.39% 134 Microbulbifer gwangyangensis 99% 0.38% 166 Vibrio sp. 99% 0.22% 167 Meridianimaribacter flavus 100% 0.22% 172 Clostridium intestinale 100% 0.21% 222 Cyclobacterium caenipelagi 100% 0.13% 224 Paenibacillus pinisoli 99% 0.13%

Table 9. Alpha diversity measures for all substrates by day sampled.

Observed Inv. Substrate Chao1 Shannon ASVs Simpson Day 7 Chi 65 65.3 ± 0.92 2.02 3.11 PE 102 104.1 ± 2.33 2.65 5.60 PVC 114 124.0 ± 6.66 2.55 7.69 PHB 121 132.0 ± 8.48 2.81 7.42 PLA 123 128.08 ± 4.04 2.67 7.69 Day 14 Chi 78 78.3 ± 0.92 2.90 8.53 PE 90 90.5 ± 1.03 2.53 5.99 PVC 67 72.5 ± 4.34 1.84 3.89 PHB 163 198.0 ± 15.71 2.88 9.15 PLA 86 87.4 ± 1.74 2.31 5.81 Day 28 Chi 176 183.0 ± 4.91 3.12 10.67 PE 90 91.5 ± 2.23 3.01 11.01 PVC 235 261.9 ± 11.19 2.75 5.72 PHB 116 121.0 ± 4.13 2.29 4.73 PLA 189 195.2 ± 4.38 3.24 12.80

Table 10. ANOVA for alpha diversity measures for all substrates.

Sum of Subset Factor Parameter Df F p Squares All Substrate Observed 4 2365 1.386 0.264 Chao 4 6314 1.847 0.148 Shannon 4 0.263 0.389 0.815 InvSimpson 4 8.99 0.377 0.823 Day Observed 2 4618 5.411 0.010 * Chao 2 8141 4.763 0.017 * Shannon 2 0.623 1.840 0.177 InvSimpson 2 27.56 2.310 0.118 Treatment Observed 2 129 0.102 0.903 Chao 2 43 0.017 0.984 Shannon 2 0.23 0.560 0.577 InvSimpson 2 6.01 0.486 0.620 Day Observed 2 4811 3.793 0.033 * Chao 2 8564 3.324 0.048 * Shannon 2 0.636 1.549 0.227 InvSimpson 2 28.05 2.265 0.119 Day 7 Substrate Observed 4 1320 2.658 0.096 Chao 4 1671 1.752 0.215 Shannon 4 0.773 2.194 0.143 InvSimpson 4 24.19 2.637 0.097 Treatment Observed 2 1260 5.805 0.017 * Chao 2 1590 3.867 0.051 Shannon 2 0.7329 4.774 0.030 * InvSimpson 2 17.41 3.515 0.063 Day 14 Substrate Observed 4 2944 3.388 0.067 Chao 4 3620 2.607 0.116 Shannon 4 2.215 5.938 0.016 * InvSimpson 4 36.29 7.269 0.009 * Treatment Observed 2 636 0.785 0.482 Chao 2 704 0.618 0.558 Shannon 2 1.24 3.604 0.066 InvSimpson 2 25.18 5.968 0.020 *

Table 10 cont.

Sum of Subset Factor Parameter Df F p Squares Day 28 Substrate Observed 4 8253 2.296 0.131 Chao 4 19069 2.540 0.106 Shannon 4 2.001 1.605 0.247 InvSimpson 4 44.06 0.821 0.540 Treatment Observed 2 987 0.365 0.702 Chao 2 2199 0.370 0.698 Shannon 2 0.783 1.084 0.369 InvSimpson 2 18.5 0.695 0.518 * significant at α = 0.05

Table 11. Tukey’s HSD test results for alpha diversity measures with significant p value from Table 8.

p Subset Factor Comparison Observed Inv. Chao Shannon ASVs Simpson All Date Day 7 - 14 0.959 0.981 - - Day 7 - 28 0.066 0.078 - - Day 14 - 28 0.043 * 0.062 - - Day 7 Treatment Control - Bio 0.014 * - 0.034 * - Petro - Bio 0.202 - 0.983 - Petro - Control 0.187 - 0.044 * - Day 14 Substrate Chi - PE - - 0.072 0.026 * Chi - PVC - - 0.065 0.024 * Chi - PHB - - 0.900 0.518 Chi - PLA - - 0.034 * 0.017 * PE - PVC - - 0.995 0.990 PE - PHB - - 0.162 0.166 PE - PLA - - 0.968 0.995 PVC - PHB - - 0.140 0.130 PVC - PLA - - 1.000 1.000 PHB - PLA - - 0.070 0.100 Treatment Control - Bio - - - 0.069 Petro - Bio - - - 0.402 Petro - Control - - - 0.016 * * statistically significant at the α = 0.05 level; - indicates not tested

Table 12a. PERMANOVA table for all days sampled (7 and 28).

Sum of Pseudo Subset Factor Df nPerm p(perm) squares F All ~Substrate 4 2.20 2.08 999 0.001 * ~Date 1 0.81 3.08 999 0.001 * ~Date:Substrate 4 1.80 1.70 999 0.001 * Treatment Bio+Petro ~Treatment 1 0.45 1.49 999 0.024 * ~Date 1 0.94 3.14 999 0.001 * ~Date:Treatment 1 0.32 1.07 999 0.319 Chi + Bio ~Treatment 1 0.74 2.55 999 0.001 * ~Date 1 0.50 1.74 999 0.007 * ~Date:Treatment 1 0.62 2.16 999 0.001 * Chi+Petro ~Treatment 1 0.75 2.60 999 0.001 * ~Date 1 0.61 2.12 999 0.001 * ~Date:Treatment 1 0.58 1.99 999 0.002 * Bio ~Substrate 1 0.40 1.41 999 0.050 * ~Date 1 0.60 2.11 999 0.001 * ~Date:Substrate 1 0.32 1.13 999 0.246 Petro ~Substrate 1 0.55 2.23 999 0.002 * ~Date 1 0.67 2.71 999 0.001 * ~Date:Substrate 1 0.50 2.05 999 0.005 *

Table 12b. PERMANOVA table subset by Day sampled.

Sum of Pseud Subset Factor Df nPerm p(perm) squares o F Day 7 All ~Substrate 4 2.18 2.47 999 0.001 * Bio+Petro ~Treatment 1 0.38 1.43 999 0.038 * Chi + Bio ~Treatment 1 0.83 3.29 999 0.011 * Chi+Petro ~Treatment 1 0.83 3.23 999 0.011 * Bio ~Substrate 1 0.41 1.84 719 0.100 Petro ~Substrate 1 0.47 2.23 719 0.100 Day 28 All ~Substrate 4 1.82 1.48 999 0.001 * Bio+Petro ~Treatment 1 0.40 1.16 999 0.169

Chi + Bio ~Treatment 1 0.52 1.62 999 0.011 * Chi+Petro ~Treatment 1 0.50 1.56 999 0.023 *

Bio ~Substrate 1 0.31 0.90 719 0.700

Petro ~Substrate 1 0.58 2.06 719 0.100 * indicates significance at the α = 0.05 level

Table 13. Permutational pairwise comparisons.

F. Subset Factor Pairs R2 nPerm p p(perm) Model All ~Substrate Chitin vs PE 2.193 0.180 999 0.003 0.027 * Chitin vs PHB 2.199 0.180 999 0.001 0.010 * Chitin vs PLA 1.921 0.161 999 0.003 0.027 * Chitin vs PVC 2.288 0.186 999 0.003 0.027 * PE vs PHB 1.490 0.130 999 0.061 0.244 PE vs PLA 1.414 0.124 999 0.063 0.244 PE vs PVC 1.746 0.149 999 0.010 0.060 PHB vs PLA 1.259 0.112 999 0.128 0.256 PHB vs PVC 1.764 0.150 999 0.010 0.060 PLA vs PVC 1.176 0.105 999 0.157 0.256 ~Date Day 7 vs Day 28 2.454 0.081 999 0.001 0.001 * ~Treatment Control vs Petro 2.297 0.126 999 0.001 0.003 * Control vs Bio 2.278 0.125 999 0.001 0.003 * Petro vs Bio 1.357 0.058 999 0.048 0.048 * Day 7 ~Treatment Control vs Petro 3.232 0.316 999 0.011 0.011 * Control vs Bio 3.294 0.320 999 0.011 0.011 * Petro vs Bio 1.433 0.125 999 0.038 0.038 * Day 28 ~Treatment Control vs Petro 1.556 0.182 999 0.024 0.024 * Control vs Bio 1.620 0.188 999 0.011 0.011 * Petro vs Bio 1.164 0.104 999 0.169 0.169

Table 14a. betadisper (PERMDISPER) table for Days 7 + 28.

Sum of Pseudo Subset Factor Df nPerm p(perm) squares F All ~Substrate 4 0.008 0.52 999 0.742

~Date 1 0.009 2.79 999 0.095

~Date:Substrate 9 0.075 0.60 999 0.758 Treatment Bio+Petro ~Treatment 1 2.71E-04 0.11 999 0.744

~Date 1 0.025 17.45 999 0.001 *

~Date:Treatment 3 0.030 4.34 999 0.017 * Chi + Bio ~Treatment 1 0.001 0.20 999 0.636

~Date 1 0.002 0.67 999 0.413

~Date:Treatment 3 0.029 0.92 999 0.480 Chi+Petro ~Treatment 1 0.001 0.50 999 0.504

~Date 1 0.001 0.52 999 0.477

~Date:Treatment 3 0.029 1.07 999 0.439 Bio ~Substrate 1 1.19E-04 0.02 999 0.876

~Date 1 0.016 4.62 999 0.073

~Date:Substrate 3 0.043 2.60 999 0.128 Petro ~Substrate 1 0.008 3.41 999 0.090

~Date 1 0.014 11.77 999 0.003 *

~Date:Substrate 3 0.017 0.38 999 0.748

Table 14b. betadisper (PERMDISPER) table subset by Day.

Sum of Subset Factor Df Pseudo F nPerm p(perm) squares Day 7 All ~Substrate 4 0.003 0.05 999 0.997

Bio+Petro ~Treatment 1 7.28E-05 0.03 999 0.837 Chi + Bio ~Treatment 1 0.003 0.19 999 0.688 Chi+Petro ~Treatment 1 0.004 0.27 999 0.592 Bio ~Substrate 1 0.001 0.20 719 0.801 Petro ~Substrate 1 0.001 0.10 719 0.701 Day 28 All ~Substrate 4 0.015 0.39 999 0.829

Bio+Petro ~Treatment 1 7.13E-06 0.003 999 0.968 Chi + Bio ~Treatment 1 0.005 0.91 999 0.327 Chi+Petro ~Treatment 1 0.005 1.47 999 0.222 Bio ~Substrate 1 3.06E-07 7.52E-05 719 0.901 Petro ~Substrate 1 0.002 0.12 719 0.801

Table 15. Putative hydrocarbonoclastic bacteria identified in the present study and percent relative abundances.

Genus Abundance Genus Abundance Exiguobacterium 19.43% Pyschrobacter 0.04% Pseudomonas 16.59% Nitratireductor 0.03% Erythrobacter 8.26% Pseudoxanthomonas 0.03% Stenotrophomonas 6.95% Gordonia 0.03% Bacillus 6.32% Sphingorhabdus 0.03% Brevundimonas 4.10% Dietzia 0.02% Paenibacillus 3.21% Microvirga 0.02% Cobetia 2.64% Arthrobacter 0.02% Lysinibacillus 2.17% Aeromicrobium 0.01% Halomonas 1.31% Sphingomonas 0.01% Streptomyces 0.82% Desulfosporosinus 0.01% Altererythrobacter 0.75% Tropicimonas 0.01% Coceicoccus 0.75% Castellaniella 0.01% Hydrogenophaga 0.64% Tenacibaculum 0.01% Lysobacter 0.38% Alteromonas 0.01% Rhodococcus 0.32% Sulfitobacter 0.01% Mycobacterium 0.30% Thalassospira 0.005% Porphyrobacter 0.26% Arenimonas 0.004% Marinobacter 0.17% Kocuria 0.004% Ruegeria 0.16% Mesorhizobium 0.004% Nocardioides 0.16% Aestuariicella 0.004% Massilia 0.13% Desulfitobacterium 0.004% Rheinheimera aquimaris 0.11% Oleiphilus 0.003% Luteimonas 0.11% Brevibacterium 0.003% Microbulbifer 0.11% Loktanella 0.003% Sphingopyxis 0.10% Sanguibacter 0.003% Ensifer 0.07% Methylobacterium 0.003% Algoriphagus 0.07% Alcanivorax 0.002% Acinetobacter 0.05% Isoptericola 0.002% Cellulosimicrobium 0.04% Flavobacterium 0.001% Brevibacillus 0.04% Total 76.85%

100

Table 16. Putative pathogenic bacteria identified in the present study and percent relative abundances.

Genus Abundance Pseudomonas 16.59% Stenotrophomonas 6.95% Bacillus 6.32% Planococcus 5.41% Halomonas 1.31% Microbacterium 0.88% Clostridium 0.42% Rhodococcus 0.32% Mycobacterium 0.30% Vibrio 0.27% Acinetobacter 0.05% Tenacibaculum 0.01% Total 38.8%

101

7. Figures

Figure 1. Comparison of taxonomic abundances and biofilm community structures of five substrates

(Chi, PE, PVC, PHB, PLA) according to immersion times (1 – Day 7; 2 – Day 14; 3 – Day 28), by

complete linkage cluster analysis based on Jaccard distances between sequencing profiles (right)

and cumulative abundance bar charts (left).

102

Figure 2. Venn diagram showing the cumulative number of shared ASVs between 5 substrates (Chi, PE, PVC, PHB, PLA)

103

Figure 3. Mean alpha diversity of the 5 substrates (Chi, PE, PVC, PHB, PLA) over each immersion time.

104

Figure 4. Principle coordinate analysis of the 5 substrates (Chi, PE, PVC, PHB, PLA) over each immersion time, showing clear clustering of biofilm communities of Day 7, Day 14, and Day 28.

Note chitin forming its own grouping in the lower left corner, and the progression of PHB from day 14 into the Day 28 group.

105

Figure 5. Betadispersion plot of Day7 and Day 28 samples from all 5 substrates (Chi, PE, PVC, PHB,

PLA). Chitin is clearly driving differences between community composition on the different substrates during Day 7 as detected by PERMANOVA.

106

8. References

Abbasian, F. et al., 2015. A comprehensive review of aliphatic hydrocarbon biodegradation by bacteria. Applied Biochemistry and Biotechnology, 176(3), pp.670–699.

Aeschelmann, F. & Carus, M., 2016. Bio-based Building Blocks and Polymers. Global Capacities and Trends 2016 – 2021, Hürth.

Altschul, S.F. et al., 1990. Basic local alignment search tool. Journal of Molecular Biology, 215(3), pp.403–410. Available at: http://linkinghub.elsevier.com/retrieve/pii/S0022283605803602.

Anil Kumar, P.K. & Suresh, P. V., 2014. Biodegradation of shrimp biowaste by marine Exiguobacterium sp. Cfr26m and concomitant production of extracellular protease and antioxidant materials: Production and process optimization by response surface methodology. Marine Biotechnology, 16(2), pp.202–218. Available at: http://link.springer.com/10.1007/s10126-013-9531-2.

Arbizu, P.M., 2017. pairwiseAdonis: Pairwise Multilevel Comparison using Adonis. R package version 0.0.1.

Artham, T. et al., 2009. Biofouling and stability of synthetic polymers in sea water. International Biodeterioration & Biodegradation, 63(7), pp.884–890. Available at: http://www.sciencedirect.com/science/article/pii/S0964830509000420 [Accessed May 23, 2016].

Brooijmans, R.J.W., Pastink, M.I. & Siezen, R.J., 2009. Hydrocarbon-degrading bacteria: The oil-spill clean-up crew. Microbial Biotechnology, 2(6), pp.587–594. Available at: http://doi.wiley.com/10.1111/j.1751-7915.2009.00151.x.

Bryant, J.A. et al., 2016. Diversity and activity of communities inhabiting plastic debris in the North Pacific gyre. mSystems, 1(3), pp.1–19. Available at: http://msystems.asm.org/lookup/doi/10.1128/mSystems.00024-16.

Callahan, B.J. et al., 2016. DADA2: High-resolution sample inference from Illumina amplicon data. Nature Methods, 13(7), pp.581–583. Available at: http://dx.doi.org/10.1038/nmeth.3869.

107

Carpenter, E.J. & Smith Jr, K.L., 1972. Plastics on the Sargasso Sea surface. Science, 175(4027), p.1240. Available at: http://www.jstor.org/stable/1733709.

Carson, H.S. et al., 2013. The plastic-associated microorganisms of the North Pacific Gyre. Marine Pollution Bulletin, 75(1–2), pp.126–132. Available at: http://dx.doi.org/10.1016/j.marpolbul.2013.07.054.

Chauhan, D. et al., 2018. Biofilm formation by Exiguobacterium sp. DR11 and DR14 alter polystyrene surface properties and initiate biodegradation. RSC Advances, 8(66), pp.37590–37599. Available at: http://xlink.rsc.org/?DOI=C8RA06448B.

Datta, M.S. et al., 2016. Microbial interactions lead to rapid micro-scale successions on model marine particles. Nature Communications, 7(1), p.11965. Available at: http://dx.doi.org/10.1038/ncomms11965.

De Tender, C. et al., 2017. Temporal dynamics of bacterial and fungal colonization on plastic debris in the North Sea. Environmental Science & Technology, 51(13), pp.7350–7360. Available at: http://pubs.acs.org/doi/abs/10.1021/acs.est.7b00697.

Debroas, D., Mone, A. & Ter Halle, A., 2017. Plastics in the North Atlantic garbage patch: A boat-microbe for hitchhikers and plastic degraders. Science of The Total Environment, 599–600, pp.1222–1232. Available at: http://dx.doi.org/10.1016/j.scitotenv.2017.05.059.

Dusa, A., 2018. venn: Draw Venn Diagrams. R package version 1.7.

Dussud, C. et al., 2018a. Evidence of niche partitioning among bacteria living on plastics, organic particles and surrounding seawaters. Environmental Pollution, 236, pp.807– 816. Available at: https://linkinghub.elsevier.com/retrieve/pii/S026974911732198X.

Dussud, C. et al., 2018b. Colonization of non-biodegradable and biodegradable plastics by marine microorganisms. Frontiers in Microbiology, 9(JUL), pp.1–13. Available at: https://www.frontiersin.org/article/10.3389/fmicb.2018.01571/full.

Elifantz, H. et al., 2013. Rhodobacteraceae are the key members of the microbial community of the initial biofilm formed in Eastern Mediterranean coastal seawater. FEMS Microbiology Ecology, 85(2), pp.348–357. Available at: https://academic.oup.com/femsec/article-lookup/doi/10.1111/1574-6941.12122.

Eren, A.M. et al., 2013. A filtering method to generate high quality short reads using Illumina paired-end technology. PLoS ONE, 8(6), p.e66643. Available at: http://dx.plos.org/10.1371/journal.pone.0066643.

Frank, J.A. et al., 2008. Critical evaluation of two primers commonly used for amplification of bacterial 16s rRNA genes. Applied and Environmental

108

Microbiology, 74(8), pp.2461–2470. Available at: http://aem.asm.org/cgi/doi/10.1128/AEM.02272-07.

Frère, L. et al., 2018. Microplastic bacterial communities in the Bay of Brest: Influence of polymer type and size. Environmental Pollution, 242, pp.614–625.

Ganesh Kumar, A. et al., 2014. Biodegradation of complex hydrocarbons in spent engine oil by novel bacterial consortium isolated from deep sea sediment. Bioresource Technology, 170, pp.556–564. Available at: http://dx.doi.org/10.1016/j.biortech.2014.08.008.

Harrison, J.P. et al., 2011. Interactions between microorganisms and marine microplastics: A call for research. Mar. Technol. Soc. J., 45(2), p.12.

Hoellein, T. et al., 2014. Anthropogenic litter in urban freshwater ecosystems: Distribution and microbial interactions. PLoS ONE, 9(6), p.e98485. Available at: http://dx.plos.org/10.1371/journal.pone.0098485 [Accessed August 6, 2016].

Jain, A. & Bhosle, N.B., 2009. Biochemical composition of the marine conditioning film: implications for bacterial adhesion. Biofouling, 25(1), pp.13–19. Available at: http://www.tandfonline.com/doi/abs/10.1080/08927010802411969.

Jambeck, J.R. et al., 2015. Plastic waste inputs from land into the ocean. Science, 347(6223), pp.768–771. Available at: http://www.sciencemag.org/cgi/doi/10.1126/science.1260352.

Jendrossek, D. & Handrick, R.R., 2002. Microbial degradation of polyhydroxyalkanoates. Annual Review of Microbiology, 56(1), pp.403–432. Available at: http://link.springer.com/10.1007/978-3-642-23789-8 [Accessed May 24, 2016].

Kasana, R.C. & Pandey, C.B., 2018. Exiguobacterium: An overview of a versatile genus with potential in industry and agriculture. Critical Reviews in Biotechnology, 38(1), pp.141–156. Available at: http://dx.doi.org/10.1080/07388551.2017.1312273.

Kwon, J.-H. et al., 2017. Microplastics as a vector of hydrophobic contaminants: Importance of hydrophobic additives. Integrated Environmental Assessment and Management, 13(3), pp.494–499. Available at: http://doi.wiley.com/10.1002/ieam.1906.

109

Lee, J.-W. et al., 2008. Bacterial communities in the initial stage of marine biofilm formation on artificial surfaces. J. Microbiol., 46(2), p.174. Available at: http://link.springer.com/10.1007/s12275-008-0032-3.

Lobelle, D. & Cunliffe, M., 2011. Early microbial biofilm formation on marine plastic debris. Mar. Pollut. Bull., 62(1), pp.197–200. Available at: http://dx.doi.org/10.1016/j.marpolbul.2010.10.013.

Lorite, G.S. et al., 2011. The role of conditioning film formation and surface chemical changes on Xylella fastidiosa adhesion and biofilm evolution. Journal of Colloid and Interface Science, 359(1), pp.289–295. Available at: http://dx.doi.org/10.1016/j.jcis.2011.03.066.

Machado, H. & Gram, L., 2015. The fur gene as a new phylogenetic marker for Vibrionaceae species identification. Applied and Environmental Microbiology, 81(8), pp.2745–2752. Available at: http://aem.asm.org/lookup/doi/10.1128/AEM.00058-15.

McCormick, A. et al., 2014. Microplastic is an abundant and distinct microbial habitat in an urban river. Environmental Science & Technology, 48(20), pp.11863–11871. Available at: http://pubs.acs.org/doi/abs/10.1021/es503610r.

McMurdie, P.J. & Holmes, S., 2013. phyloseq: An R package for reproducible interactive analysis and graphics of microbiome census data. PLoS ONE, 8(4), p.e61217. Available at: https://dx.plos.org/10.1371/journal.pone.0061217.

Mohanty, G. & Mukherji, S., 2008. Biodegradation rate of diesel range n-alkanes by bacterial cultures Exiguobacterium aurantiacum and Burkholderia cepacia. International Biodeterioration & Biodegradation, 61(3), pp.240–250. Available at: http://linkinghub.elsevier.com/retrieve/pii/S0964830507001321.

Muthukumar, T. et al., 2014. Biodegradation of starch blended high density polyethylene using marine bacteria associated with biofilm formation and its isolation characterization. Journal of Microbial & Biochemical Technology, 06(03), pp.116– 122. Available at: https://www.omicsonline.org/open-access/biodegradation-of- starch-blended-high-density-polyethylene-using-marine-bacteria-associated-with- biofilm-formation-and-its-isolation-characterization-1948- 5948.1000131.php?aid=23791.

110

Oberbeckmann, S., Kreikemeyer, B. & Labrenz, M., 2018. Environmental factors support the formation of specific bacterial assemblages on microplastics. Frontiers in Microbiology, 8(JAN), pp.1–12. Available at: http://journal.frontiersin.org/article/10.3389/fmicb.2017.02709/full.

Oberbeckmann, S., Löder, M.G.J. & Labrenz, M., 2015. Marine microplastic-associated biofilms – A review. Environmental Chemistry, 12(5), p.551. Available at: http://www.publish.csiro.au/view/journals/dsp_journal_fulltext.cfm?nid=188&f=EN 15069 [Accessed February 23, 2016].

Oberbeckmann, S. et al., 2014. Spatial and seasonal variation in diversity and structure of microbial biofilms on marine plastics in Northern European waters. FEMS microbiology ecology, 49(0), pp.478–492. Available at: http://www.ncbi.nlm.nih.gov/pubmed/25109340 [Accessed May 20, 2016].

Oberbeckmann, S., Osborn, A.M. & Duhaime, M.B., 2016. Microbes on a Bottle: Substrate, season and geography influence community composition of microbes colonizing marine plastic debris. PLOS ONE, 11(8), p.e0159289. Available at: http://dx.plos.org/10.1371/journal.pone.0159289.

Ogonowski, M. et al., 2018. Evidence for selective bacterial community structuring on microplastics. Environmental Microbiology, 20(8), pp.2796–2808. Available at: http://doi.wiley.com/10.1111/1462-2920.14120.

Oksanen, J. et al., 2018. vegan: Community Ecology Package. R package version 2.5-3.

Parada, A.E., Needham, D.M. & Fuhrman, J.A., 2016. Every base matters: Assessing small subunit rRNA primers for marine microbiomes with mock communities, time series and global field samples. Environmental Microbiology, 18(5), pp.1403–1414.

Quast, C. et al., 2012. The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic Acids Research, 41(D1), pp.D590–D596. Available at: http://academic.oup.com/nar/article/41/D1/D590/1069277/The- SILVA-ribosomal-RNA-gene-database-project.

Quince, C. et al., 2011. Removing noise from pyrosequenced amplicons. BMC Bioinformatics, 12(1), p.38. Available at: http://bmcbioinformatics.biomedcentral.com/articles/10.1186/1471-2105-12-38.

Reen, F.J. et al., 2006. The genomic code: inferring Vibrionaceae niche specialization. Nature Reviews Microbiology, 4(9), pp.697–704. Available at: http://www.nature.com/articles/nrmicro1476.

111

Schultz, G., White, E. & Ducklow, H., 2003. Bacterioplankton dynamics in the York River estuary: primary influence of temperature and freshwater inputs. Aquatic Microbial Ecology, 30(2), pp.135–148. Available at: http://www.int- res.com/abstracts/ame/v30/n2/p135-148/.

Sigler, M., 2014. The effects of plastic pollution on aquatic wildlife: Current situations and future solutions. Water, Air, & Soil Pollution, 225(11), p.2184. Available at: http://link.springer.com/10.1007/s11270-014-2184-6.

Stults, J.R. et al., 2001. Application of the 5’ fluorogenic exonuclease assay (TaqMan) for quantitative ribosomal DNA and rRNA analysis in sediments. Applied and Environmental Microbiology, 67(6), pp.2781–2789. Available at: http://dblp.uni- trier.de/db/conf/acss/acss2014-1.html.

Sudhakar, M. et al., 2007. Biofouling and biodegradation of polyolefins in ocean waters. Polymer Degradation and Stability, 92(9), pp.1743–1752. Available at: http://www.sciencedirect.com/science/article/pii/S0141391007001267 [Accessed May 24, 2016].

Sweet, M.J., Croquer, A. & Bythell, J.C., 2011. Development of bacterial biofilms on artificial corals in comparison to surface-associated microbes of hard corals. PLoS ONE, 6(6), p.e21195. Available at: https://dx.plos.org/10.1371/journal.pone.0021195.

R: A language and environment for statistical computing. Available at: https://www.r- project.org.

Tonon, L.A.C., Moreira, A.P.B. & Thompson, F., 2014. The Family Erythrobacteraceae. In E. Rosenberg et al., eds. The Prokaryotes. Berlin, Heidelberg: Springer Berlin Heidelberg, pp. 213–235. Available at: https://doi.org/10.1007/978-3-642-30197- 1_376.

Whitaker, D. & Christman, M., 2014. clustsig: Significant Cluster Analysis. R package version 1.1.

Yang, J. et al., 2014. Evidence of polyethylene biodegradation by bacterial strains from the guts of plastic-eating waxworms. Environmental Science & Technology, 48(23), pp.13776–13784. Available at: http://pubs.acs.org/doi/abs/10.1021/es504038a [Accessed May 24, 2016].

112

Ye, S. & Andrady, A.L., 1991. Fouling of floating plastic debris under Biscayne Bay exposure conditions. Marine Pollution Bulletin, 22(12), pp.608–613. Available at: http://linkinghub.elsevier.com/retrieve/pii/0025326X9190249R.

Zettler, E.R., Mincer, T.J. & Amaral-Zettler, L.A., 2013. Life in the “Plastisphere”: Microbial communities on plastic marine debris. Environmental Science & Technology, 47(13), p.130619162220002. Available at: http://dx.doi.org/10.1021/es401288x [Accessed December 21, 2015].

Zhang, H. et al., 2015. Biodegradation of polychlorinated biphenyls (PCBs) by the novel identified cyanobacterium Anabaena PD-1. PLoS ONE, 10(7), pp.1–16.

113

POLICY CONSIDERATIONS

1. Introduction With public attention on marine debris and microplastic pollution, there is increased pressure on regulatory bodies to take action to prevent and limit the amount of plastics entering the marine environment. Most federal policies concerning marine debris predates our current knowledge of the nuanced effects of marine microplastics. In the

1970’s two laws sought specifically to regulate the pollution of the oceans by foreign material. The first is The Marine Protection, Research, and Sanctuaries Act (33 U.S.C.

§§1401 et seq. (2017)), often referred to as The Ocean Dumping Act. This Act regulates the transport and dumping of material into ocean waters. The second is the Act to Prevent

Pollution from Ships (APPS, 33 U.S.C §§1901 et seq. (2017)). APPS implements the

International Convention for the Prevention of Pollution from Ships (MARPOL) annexes

I – IV in U.S. waters. In 1987, APPS was amended to cover MARPOL Annex V, known as the Marine Plastic Pollution Research and Control Act. This Act sought to eliminate and reduce garbage discharge (including all plastics) from U.S. ships around the world and foreign ships navigating U.S. waters, but excluded ships of the Armed Forces. This gap was closed with an amendment in 2011 (33 U.S.C. §1902 (2017)), which specifically targeted the discharge of plastics and called for the development of technologies to reduce waste streams from Armed Forces ships.

These two acts have certainly not eliminated the introduction of plastics into the marine environment but have cut down on intentional dumping. More recent laws include the Marine Debris Act and the Microbead-Free Waters Act. In 2006, the Marine Debris

Act (MDA, 33 U.S.C. §§1951-1958 (2017)) established NOAA’s Marine Debris Program

114

to identify, prevent, reduce, and remove marine debris and address the adverse impacts of marine debris on the marine environment. A 2012 amendment of MDA defined marine debris to include “any persistent solid material…disposed of or abandoned into the marine environment or Great Lakes.” NOAA’s Marine Debris Program Marine Debris

Clearinghouse currently references 54 projects related to microplastic debris (2019), ranging from investigating ecotoxicological effects on fishes to sponsoring community- outreach and debris removal. In 2015, the Microbead-Free Waters Act (21 U.S.C.

§331(ddd) (2017)) amended the Federal Food, Drug, and Cosmetic Act to prohibit intentionally added plastics microbeads in commercial rinse-off products, such as face wash and toothpaste. This was a way to remove some manufactured microplastics from entering wastewater streams that ultimately dump into waterways. A majority of microplastics are secondary in nature, meaning that they are created from the breakdown of larger plastic material, so this Law does little to address the larger issue at hand.

The most recent Act relevant to marine plastic debris is the 2018 Save Our Seas Act

(P.L. No. 115-265), an amendment of the Marine Debris Act. It furthers the mission of

NOAA’s Marine Debris Program to develop outreach and education strategies to address land-based sources of marine debris and to promote international action to reduce the incidence of marine debris. Additionally, §§102 and 103 of the Act are “Sense of

Congress” statements that, while are not enforceable or binding, indicate shifting priorities within the legislature. Federal agencies and foreign governments take note of these “Sense of” resolutions as they can indicate what Congress supports or opposes, how funding decisions may be made, and U.S. foreign affair priorities (Longley, 2018). In the

Save Our Seas Act, Section 102 gives a Sense of Congress on International Engagement

115

to Respond to Marine Debris, specifically expressing supporting research and development on systems and materials that reduce derelict fishing gear and solid waste from land-based sources that enters the marine environment. This is directly applicable to my research as I focused on two bio-based plastics that have been gaining traction in manufacturing sectors as benign alternatives to conventional petroleum-based plastics.

2. Considerations for bio-based plastics

The advantages of bio-based, biodegradable materials lie in the feedstock sourcing and the product end-of-life considerations. Bio-based plastics are produced from renewable carbon sources, improving resource security, as current production of plastics utilize about 8% of the global crude oil and gas production (Lambert & Wagner 2017).

There is some concern over the overall environmental impact of bio-based materials, but it is beyond the scope of this paper to examine life cycle assessments (LCAs) of PHA and

PLA (for a recent review of LCAs, see Yates & Barlow 2013). The motivation for promoting development of bio-based products lies more with the end-of-life, especially for products that have a short lifetime or are easily lost into the environment. Currently,

PLA is the more popular and widely used bio-based plastic, but it is not truly biodegradable. Instead, PLA is considered compostable in industrial composting facilities, where it can lose 34% weight after 7 weeks at 65℃ (Lambert & Wagner 2017).

PHA on the other hand is wholly biodegradable, as it is produced by a wide-range of bacteria and functions as intercellular carbon stores. This means that a wide-range of bacteria are also capable of breaking it down. When inadvertently disposed of into the marine environment, PLA is not likely to biodegrade (Pelegrini et al. 2016), but PHA has been demonstrated to degrade under marine conditions (Volova et al. 2010).

116

When making sweeping policy changes, it is important for science advisors and policymakers to consider the differences in alternative plastics. My work provides evidence to support the use of bio-based materials in single-use plastic products. The bio- based plastics PLA and PHA are not as sorptive of HOCs as PE (especially PLA; this work), which decreases the potential transport of sorbed contaminants into the food web.

PLA and PHA products would certainly still generate microplastics, so it is important to consider their impacts if ingested by marine biota. Further work is required to examine the toxicity of these plastics on marine indicator species.

3. Future considerations

With any plastic materials, additives are used to enhance or expand chemical properties. These additives are often present in high-levels within a final plastic product and have the ability to leach out into the water column or into the gut fluid of ingesting organisms. With bio-based materials that may be marketed as ‘eco-friendly,’ the scrutiny of additives must be thorough. McDevitt et al. (2017) define the term ‘ecocyclable’ as a material that satisfies certain standards for degradability, bioaccumulation, and toxicity.

These standards include that the material, as well as the associated additives, do not bioaccumulate and that toxicity under acute and chronic exposure is not significantly greater than benign reference materials. Along with cellulose, raw poly-3- hydroxybutyrate is one of these reference materials for the ecocyclable definition, due to its natural occurrence, biodegradability, and non-toxic degradation products (McDevitt et al. 2017). This definition should be considered with regards to future legislation concerning new technologies to address marine debris, and in particular marine plastic debris. Future needs in bio-based plastics research include investigations into aquatic

117

(bio)degradation of commercial products, additive use and toxicology, including impacts when ingested by biota and impacts to local microbial communities.

118

4. References

Lambert, S., & Wagner, M. (2017). Environmental performance of bio-based and biodegradable plastics: the road ahead. Chemical Society Reviews, 46(22), 6855– 6871. https://doi.org/10.1039/C7CS00149E

Longley, Robert. (2018). What Is a 'Sense of Congress' Resolution? Available from: https://www.thoughtco.com/sense-of-congress-resolutions-3322308

McDevitt, J. P., Criddle, C. S., Morse, M., Hale, R. C., Bott, C. B., & Rochman, C. M. (2017). Addressing the Issue of Microplastics in the Wake of the Microbead-Free Waters Act—A New Standard Can Facilitate Improved Policy. Environmental Science & Technology, 51(12), 6611–6617. https://doi.org/10.1021/acs.est.6b05812

NOAA Marine Debris Program (2019). Marine Debris Clearinghouse. Retrieved April 15, 2019, from https://clearinghouse.marinedebris.noaa.gov/

Pelegrini, K., Donazzolo, I., Brambilla, V., Coulon Grisa, A. M., Piazza, D., Zattera, A. J., & Brandalise, R. N. (2016). Degradation of PLA and PLA in composites with triacetin and buriti fiber after 600 days in a simulated marine environment. Journal of Applied Polymer Science, 133(15), 43290(1-7). https://doi.org/10.1002/app.43290

Volova, T. G., Boyandin, A. N., Vasiliev, A. D., Karpov, V. A., Prudnikova, S. V., Mishukova, O. V., … Gitelson, I. I. (2010). Biodegradation of polyhydroxyalkanoates (PHAs) in tropical coastal waters and identification of PHA- degrading bacteria. Polymer Degradation and Stability, 95(12), 2350–2359. https://doi.org/10.1016/j.polymdegradstab.2010.08.023

Yates, M. R., & Barlow, C. Y. (2013). Life cycle assessments of biodegradable, commercial biopolymers - A critical review. Resources, Conservation and Recycling, 78, 54–66. https://doi.org/10.1016/j.resconrec.2013.06.010

119

VITAE

Kelley Ann Uhlig was born in Hollywood, FL. She graduated from Williston High School in Williston, FL in the top 10 of her class. Kelley attended the University of North Florida (UNF) in Jacksonville, FL, graduating in 2013 with a Bachelor’s of Science in Chemistry. While at UNF, she received recognition for her undergraduate research work as an outstanding scholar in nano-material doped glass under Dr. José Jimenez. Her work on manganese cluster synthesis under Dr. Christos Lampropoulos earned her authorship recognition in a 2015 publication. Kelley turned her attention toward environmental chemistry and marine debris, and matriculated with the class of 2014 to the Virginia Institute of Marine Science at William & Mary under the supervision of Dr. Robert Hale. Kelley is currently completing a John A. Knauss Marine Policy Fellowship with the National Oceanographic and Atmospheric Administration’s Ocean Observation and Monitoring Division in Silver Spring, MD.

120