Scientific Significance Statement
Plastics have innumerable uses and are inextricably tied to daily life in modern society. These plastics begin as or break down into microplastics, which are now found in an array of terrestrial, aquatic, and marine habitats and organisms. These tiny particles may threaten ecosystem balance and natural resource consumers, particularly in the case of seafood. In Oregon, U.S.A., Pacific oysters (Crassostrea gigas) and razor clams (Siliqua patula) are of commercial, recreational, and cultural importance, yet baseline information on microplastic prevalence in these species and across sites and seasons is absent. Our study is the first to document microplastics in Pacific razor clams and provides important coast‐wide data to compare microplastic burden across species, seasons, and sites.
Microplastics, plastics 0.0001–5 mm in any linear direction (UNEP 2016), are found in nearly every environment on earth (Thompson et al. 2004). These tiny fragments, pellets, filaments, and fibers originate from both marine and land‐based sources, infiltrating aquatic ecosystems worldwide through pollution, runoff, wastewater, and atmospheric deposition (Zhang 2017). Globally, the overwhelming number of single‐use and nondegradable plastic items has led to widespread microplastic pollution. Plastics are manufactured to be durable, so degradation can take hundreds to thousands of years, posing a pervasive and severe problem for ecosystems as well as a human health concern (Cole et al. 2011; Wang et al. 2016).
While spatial distribution of microplastics in the environment is highly complex, areas with high human population, coastal recreation, and tourism pressures generally yield high environmental microplastics (Barnes et al. 2009; Hantoro et al. 2019). Microplastics represent a diverse set of contaminants which encompass infinite combinations of plastic densities, sizes, shapes, surface textures, and chemical properties (Rochman et al. 2019). Once transmitted into the environment, microplastics are subjected to an array of dynamic hydrological, biological, and atmospheric processes, including surface currents, tides, biofouling, mechanical and ultraviolet degradation, precipitation, storm events, and more. While human presence may correlate with microplastic prevalence, it is unclear what specific environmental processes best predict fate and transport of these pernicious particles (Jambeck et al. 2015; Zhang 2017). Density has been thought to ultimately determine environmental fate, with denser plastics like polyvinyl chloride and polyethylene terephthalate (PET) settling to the benthos and low density polymers such as polystyrene, polypropylene, and polyethylene remaining in surface waters; however, a recent review of global surface water and sediment data indicates a mixture of high and low density microplastics in water and sediment samples, attributed to influences of varied environmental and biological processes in coastal areas (Hantoro et al. 2019).
Rivers have been well established as vectors of plastics into coastal and marine environments (Zhang 2017). These dynamic waterways transport between 1.2 and 2.4 million tons of plastic into global oceans each year, with up to 28.8 thousand tons transmitted annually through North and Central American rivers alone (Lebreton et al. 2017). A study investigating microplastic concentrations in surface waters from two Los Angeles, California rivers quantified an input of roughly 2 billion microplastic pieces into coastal waters in the span of just 3 d (Moore et al. 2005). Expanded to an annual output, these two rivers transport over 240 billion microplastics per year to the California coast. Stormwater runoff and wastewater treatment plant (WWTP) effluent also contribute significant microplastic burdens to coastal environments (e.g., Carr et al. 2016; Napper and Thompson 2016; Mintenig et al. 2017). Microfibers, generally broken down from laundered clothing items or from derelict fishing gear, are the most prevalent form of microplastic in the nearshore environment (Barrows et al. 2018; de Falco et al. 2019).
Organisms inhabiting coastal environments are subjected to ambient environmental conditions, including microplastic contamination that may exist in surrounding waters, substrata, or in the air. Aquatic filter‐ and suspension‐feeding organisms can encounter microplastics in the marine or freshwater column, mistake them for food items, and ingest them (Hantoro et al. 2019). This transfer of plastics from the environment into aquatic food webs has been documented across diverse taxonomic groups, life histories, habitats, and feeding types (e.g., Cole et al. 2011; Akpan 2014; Rochman et al. 2015; Waite et al. 2018). After uptake, microplastics can adhere to organs or become incorporated into guts, gills, and tissues of organisms, decreasing energy uptake and impairing muscle function and reproduction (e.g., von Moos et al. 2012; Sussarellu et al. 2016; Ribeiro et al. 2017; Kolandhasamy et al. 2018). Microplastics may also sorb harmful contaminants that, once ingested or incorporated in tissues, are released into the organism (Teuten et al. 2007, 2009). In some studies, environmental microplastic concentrations have been directly correlated with microplastic burdens in coastal bivalves (Mathalon and Hill 2014; Li et al. 2016; Qu et al. 2018; Hantoro et al. 2019).
As filter feeders, bivalves are particularly vulnerable to contaminants in the estuarine and open coast environments they inhabit. The Pacific Northwest (PNW) region of North America supports an array of filter‐feeding shellfish species, which have been inextricably tied to the natural history and cultural heritage of the area for millennia. Fishery and aquaculture sectors serve as important anchors of the region, with Pacific oysters and razor clams playing particularly significant roles in food security and the economy (Crossett et al. 2013). Pacific oysters (Crassostrea gigas) have been commercially farmed in the PNW since introduction of the species in the early 1900s (Glude and Chew 1982). These filter feeders consume particulates in the water column, such as plankton and other organic material, and reach commercial size (100–150 mm, maximum length 250 mm) over 2–4 years (Pauley et al. 1988; Harris 2008). Pacific razor clams (Siliqua patula) are native to the PNW and are found on intertidal beaches. They have been harvested by first nations and tribal peoples for centuries, and in state‐managed recreational and commercial fisheries since the 1950s. Consuming phytoplankton, razor clams grow rapidly in the first year attaining lengths up to 90 mm, and a maximum length of 16 cm over their 6‐yr lifespan (Link 2000).
Microplastic concentrations in field‐collected Pacific oysters have been documented worldwide facilitating comparisons between samples grown in Oregon vs. other regions; however, there is no published literature on microplastic prevalence or effects in Pacific razor clams. We initiated this study to answer the question: What variables predict microplastic concentrations in Oregon Pacific oysters and Pacific razor clams?
Flowing between the U.S.A. states of Washington and Oregon is the Columbia River, the largest river on the North American continent with a Pacific Ocean terminus. We predicted the Columbia would be a major vector of microplastics, causing elevated burdens in our study species at the northernmost study sites and attenuated burdens with increased distance from the river. Coastal tourism is highest during the summer months (May–October). Tourism results in increased use of beaches and waterways for recreation and an uptick in laundering needs, so we hypothesized that concentration of environmental microplastics in waters and coastal organisms would be higher in summer than spring. We hypothesized that gut tissue would contain more microplastics than nongut tissue due to retention of microplastics in the gut of bivalves observed in previous studies (e.g., Browne et al. 2008; Ward and Kach 2009; Sussarellu et al. 2016; Woods et al. 2018). Because microplastics may become lodged in gills or other organs (Woods et al. 2018), we predicted a positive relationship between organism size and microplastic burden—that larger individuals would contain more microplastics than smaller individuals. We examined these expectations through field‐collection of Pacific razor clams and purchase of Pacific oysters at 15 locations, during two seasons, taking biological measurements and investigating whole, gut‐tissue, and tissue‐only samples.
Field sites, sample collection, processing, and microplastic enumeration
A total of 141 Pacific oysters and 142 Pacific razor clams were collected from 15 sites during low tides in spring (27–28 April 2017) and summer (21–31 July 2017) (Table 1). Whole oysters were purchased from growers at six sites during both seasons. One oyster grower was selected from each of six Pacific oyster‐producing bays. In this report, oyster grower names are withheld and are coded randomly as OY1–OY6. Oyster shell length averaged 125.39 mm (range = 77.67–197.66 mm) and wet tissue weight averaged 30.97 g (range = 8.51–101.67 g; Supporting Information Appendix 1).
|# samples analyzed||Microplastic burden||# samples analyzed||Microplastic burden|
|Species||Site||Whole organisms||Gut tissue||Nongut tissue||Avg # MP per sample (SE)a||Avg # MP g−1 tissue (SE)b||Whole organisms||Gut tissue||Nongut tissue||Avg # MP per sample (SE)a||Avg # MP g−1 tissue (SE)b||Avg MP length in mm (SE)c||Min MP length (mm)d||Max MP length (mm)e|
|Pacific oyster||OY1||10||0||0||13.60 (2.60)||0.55 (0.34)||10||3||3||9.6 (2.56)||0.49 (0.17)||1.10 (0.05)||0.16||5.37|
|OY2||12||0||0||10.33 (1.92)||0.35 (0.29)||11||3||3||6.81 (1.58)||0.21 (0.05)||1.32 (0.06)||0.18||5.85|
|OY3||10||0||0||14.60 (3.53)||0.62 (0.49)||10||2||3||8.5 (2.13)||0.28 (0.07)||1.24 (0.07)||0.12||6.08|
|OY4||10||0||0||17.50 (3.85)||0.39 (0.28)||10||3||3||5.20 (1.54)||0.10 (0.02)||1.23 (0.05)||0.11||5.42|
|OY5||10||0||0||16.30 (2.80)||0.85 (0.41)||10||3||2||7.7 (1.48)||0.57 (0.16)||1.24 (0.05)||0.10||5.56|
|OY6||10||0||0||10.80 (2.01)||0.31 (0.16)||11||3||2||11.00 (3.03)||0.50 (0.17)||1.31 (0.07)||0.15||5.40|
|Pacific razor clam||Clatsop Beach||10||0||0||9.50 (1.21)||0.18 (0.09)||10||3||3||7.60 (1.01)||0.13 (0.02)||1.30 (0.07)||0.19||5.04|
|Cannon Beach||5||0||0||10.00 (1.48)||0.18 (0.05)||9||3||3||9.78 (1.47)||0.17 (0.02)||1.43 (0.09)||0.18||8.19|
|Cape Meares||10||0||0||8.00 (2.82)||0.19 (0.22)||7||3||3||7.00 (2.57)||0.62 (0.33)||1.19 (0.07)||0.16||4.27|
|Agate Beach||0||0||0||N/A||N/A||10||3||3||6.3 (0.91)||0.09 (0.01)||1.32 (0.12)||0.26||5.73|
|Newport S. Beach||13||0||0||9.69 (1.49)||0.21 (0.12)||10||3||3||9.30 (0.84)||0.14 (0.02)||1.44 (0.07)||0.21||7.04|
|Coos Bay||12||0||0||10.50 (1.55)||0.23 (0.10)||0||0||0||N/A||N/A||1.46 (0.08)||0.31||4.73|
|Bastendorff Beach||0||0||0||N/A||N/A||5||0||0||14.80 (1.24)||0.25 (0.01)||1.38 (0.10)||0.27||6.09|
|Whiskey Creek||0||0||0||N/A||N/A||10||3||3||6.30 (0.87)||0.11 (0.02)||1.54 (0.13)||0.36||5.71|
|Gold Beach||0||0||0||N/A||N/A||10||3||3||8.70 (1.47)||0.12 (0.02)||1.29 (0.09)||0.26||4.93|
- Notes: Avg., Average; MP, microplastic; SE, ± standard error; OY1–OY6: oyster site (randomized). Reported values include background and processing fiber levels.
- aAverage number of MP per sample (SE).
- bAverage number of MP per gram of tissue (SE).
- cAverage MP length in millimeters (SE).
- dMinimum MP length at site in millimeters (SE).
- eMinimum MP length at site in millimeters (SE).
Razor clams were collected from nine sandy beach sites stretching from Clatsop in the north, to Gold beach, near the California border, in the south (Fig. 1). Of the nine clam sites, four were sampled in both spring and summer, providing a temporal snapshot of microplastic frequencies. Collection was performed in coordination with Oregon Department of Fish and Wildlife (ODFW) and Oregon Department of Agriculture (ODA), which greatly augmented efforts to achieve desired sample size. Clam sites were selected based on ODFW knowledge of existing clam populations, feasibility of sample collection (access, tides, clam shows), and with a goal of sampling a large swath of the coast. Summer clam sampling was more robust than spring because it corresponded with a coast‐wide ODFW survey and coincided with lower tides than spring. Razor clam shell length averaged 113.89 mm (range = 56.00–132.52 mm) and wet tissue weight averaged 55.71 g (range = 5.84–92.11 g; Supporting Information Appendix 1).
All samples were transported on ice to the Applied Coastal Ecology laboratory at Portland State University (PSU) in Portland, Oregon, in clean 2‐L glass Mason jars. Shell and tissue measurements were collected with a digital Mitutoyo caliper and Ohaus balance accurate to 0.01 mm and 0.01 g, respectively. Bivalve shells were rinsed with deionized (DI) water to remove sand, mud, and debris, were shucked into clean 120 mL Mason jars and frozen at −20°C.
Samples were thawed and digested for 24 h in a laminar flow fume hood using 10% potassium hydroxide (KOH). Digestion began with the first organism from each site and season, then proceeded to the second organism from each site, until all samples were processed. Samples were poured through a 7.6 cm diameter, 63 μm stainless steel sieve. Material retained on the sieve was rinsed into clean, labeled glass petri dishes. Petri dishes with Petristickers® affixed to the bases were placed in a drying oven at 40°C for 24 h and stored in sealed tubs prior to microscope processing. Due to high levels of organic material and sand granules remaining in clam samples after initial digestion, a second 10% KOH digestion combined with hypersaline density separation (330 g L−1 Fisher Chemical Certified ACS Crystalline NaCl) was utilized. Samples were analyzed under a Leica M165C stereomicroscope (×10–120 magnification) connected via a Leica IC80 HD camera to a computer running Leica Application Suite X imaging software. Each suspected microplastic encountered was measured and particle category (fiber, fragment, film, foam, bead, unknown), color, and maximum length were recorded. To determine material type for microplastics, a subset of identified fibers was randomly selected using random number generation to determine: (1) sample dish, then (2) segment of each dish (segment numbers 1–16) from which to extract 26 suspected microplastics. The first fiber visually encountered in the randomly generated dish and segment was selected for validation. Fibers were analyzed using a Thermo Nicolet iS10 Fourier‐transform infrared spectrometer (FTIR) equipped with an Attenuated Total Reflectance accessory at the University of New Hampshire Instrumentation Center. Spectra for each microfiber were acquired using 256–1024 scans depending on size and width. Automatic software comparison of microfiber spectra to a set of Thermo Nicolet Omnic™ FTIR spectral libraries was used to generate a best match.
During summer sampling, three individual organisms from each site (with the exceptions of Bastendorff Beach and Coos Bay) underwent a separation of digestive organs from other tissues. For Pacific oysters, gut‐tissue samples included the visceral mass, esophagus, diverticular gland, midgut, and stomach. In razor clams, gut‐tissue samples included the stomach, small intestine, and crystalline style. All remaining tissue was classified as nongut tissue. Separated gut and nongut tissues underwent the same digestion and microscope analyses as whole organism samples.
Quality control: Contamination quantification and prevention
One hundred percent cotton clothing, cotton lab coats, and nitrile gloves were worn at all times during sample processing, digestion, and analysis procedures. All shucking implements and glassware were rinsed three times with DI water filtered to 0.22 μm. To quantify procedural contamination, 11 replicates of 50 mL filtered DI water were frozen in 4 oz jars and underwent the same digestion and analysis process as organism samples. One procedural blank per week was chemically digested alongside field samples on a randomly generated day. Additionally, three procedural blanks were collected to quantify contamination introduced by the secondary digestion and hypersaline density separation of razor clam samples.
During microscope analysis, a petri dish containing filtered DI water was placed adjacent to each sample on the microscope base and left open to the air to quantify airborne contaminants. After sample analysis, the control petri dish was analyzed for microplastics; any particles detected were assumed to be contamination and were measured and categorized.
Data analysis and availability
To identify differences between sample sites, seasons, and anatomical burdens, ANOVA and Welch’s t‐tests were conducted in the R statistical program (v1.2.1335) using the aov and t.test functions (R Core Team 2019). Linear regression models were used to examine relationships between biological parameters (shell length, body weight) and microplastic burdens. Microplastic concentrations are expressed as number of microplastics per sample or mean number of particles g−1 tissue (wet weight; whole organisms only). Number of microplastics per sample and number of microplastics per gram of tissue variables were log transformed (log x 1) prior to statistical analysis. The statistical cutoff (alpha) for all tests was 0.05 with standard error (SE) reported. Data and metadata are available in the Portland State University PDXScholar data repository.
Numerous measures were taken to minimize procedural contamination, but as with other studies (e.g., Li et al. 2015; Davidson and Dudas 2016; Qu et al. 2018; Su et al. 2018) it was not completely eliminated. Contamination in procedural controls (4.91 ± 1.11), microscope blanks (0.20 ± 0.03), and, for razor clams, a secondary digestion and separation step (1.0 ± 0.0) was quantified (Supporting Information Appendix 2). From these controls and procedural blanks, total contamination in oyster and clam samples was estimated at 5.11 and 6.11 microplastics per sample, respectively. Average microplastic length detected as contamination (n = 124) for all sample types was 1.67 ± 0.11 mm, and most frequently detected colors in blanks and controls were colorless (79%) and blue (10%). As with multiple other studies (e.g., Li et al. 2015, 2016, 2018a; Davidson and Dudas 2016; Qu et al. 2018; Su et al. 2018; Rochman et al. 2019), we report microplastics detected in blank samples (Supporting Information Appendix 2), rather than performing a blank‐subtraction on environmental results since controls were intended to provide a range of possible contamination levels introduced through laboratory procedures. As such, our reported numbers are estimated maximum possible microplastic concentrations.
Microplastic occurrence in study species
A total of 3,053 suspected microplastics were isolated from 320 whole‐organism, gut‐tissue, and nongut tissue samples. Over 99% of particles were microfibers (n = 3,026) averaging 1.34 mm in length (range = 0.10–8.72 mm). The remaining < 1% of microplastics were categorized as fragments (n = 12), bea 1%i>