Microplastics in Nearshore and Subtidal Sediments in the Salish Sea: Implications for Marine Habitats and Exposure

Abstract

Plastic debris is a pervasive and persistent threat to marine ecosystems. Microplastics (plastics < 5 mm) are increasing in a variety of marine habitats, including open water systems, shorelines, and benthic sediments. It remains unclear how microplastics distribute and accumulate in marine systems and the extent to which this pollutant is accessible to marine taxa. We examined subtidal benthic sediments and beach sediments in critical nearshore habitats for forage fish species—Pacific sand lance (Ammodytes personatus), Pacific herring (Clupea pallasi), and surf smelt (Hypomesus pretiosus)—to quantify microplastic concentrations in the spawning and deep-water habitats of these fish and better understand how microplastics accumulate and distribute in nearshore systems. In the San Juan Islands, we examined an offshore subtidal bedform in a high-flow channel and beach sites of protected and exposed shorelines. We also examined 12 beach sites proximate to urban areas in Puget Sound. Microplastics were found in all samples and at all sample sites. Microfibers were the most abundant, and flakes were present proximate to major shipyards and marinas. Microplastics were significantly elevated in Puget Sound compared to the San Juan Archipelago. Protected beaches had elevated concentrations relative to exposed beaches and subtidal sediments. Microplastics were in higher concentrations in sand and fine-grain sediments, poorly sorted sediments, and artificial sediments. Microplastics were also elevated at sites confirmed as spawning habitats for forage fish. The model results indicate that both current speed and proximate urban populations influence nearshore microplastic concentrations. Our research provides new insights into how microplastics are distributed, deposited, and retained in marine sediments and shorelines, as well as insight into potential exposure in benthic, demersal, and shoreline habitats. Further analyses are required to examine the relative influence of urban populations and shipping lanes and the effects of physical processes such as wave exposure, tidal currents, and shoreline geometry.

1. Introduction

Plastics are synthetic polymers prevalent as durable, lightweight, low-cost materials, and are also a persistent pollutant. In the marine environment, naturally occurring environmental conditions and processes, particularly ocean current dynamics, solar radiation, abrasion, and interactions with organisms, cause plastic items to degrade and fragment into smaller particles (0.1 to 5000 μm), commonly known as microplastics (MPs) [1].

Although widespread, MP distribution in coastal and marine environments is uneven and difficult to predict because atmospheric, coastal, and tidal processes contribute to their movement, dispersal, and accumulation [2]. Most plastics have a lower density than seawater, which facilitates suspension in the water column and mixing in the surface boundary layer [3]. Vertical distribution is influenced by wind-induced turbulence, surface waves, and Langmuir circulation [3]. Within coastal areas, MPs are subject to alongshore transport, tidal transport, and wave transport [4,5]. Plastics may also sink to the seafloor [6,7,8]. Interaction of MPs with organic and inorganic matter via biofouling and absorption may alter densities and settling rates [4]. Coastal marine ecosystems often have higher concentrations of MP pollution compared to offshore areas [9,10,11,12]. Nearshore environments and the organisms that reside there may therefore be most directly affected.

The North Pacific Ocean is a global hotspot for MPs [13]. The Salish Sea (Figure 1) is a coastal estuarine system in the eastern North Pacific proximate to multiple urban areas where plastics and MPs concentrate [1,14] and an area of concern. In coastal areas, local densities vary due to complex fine-scale oceanographic and anthropogenic factors. Fjord estuaries, such as the Salish Sea, have complex bathymetry [15,16], including deep basins with shallow sills, resident sediments, and complex circulation dynamics [17,18]. Freshwater input from rivers also plays an important role [19]. The combination of deep basins, shallow sills, and river influence impacts circulation, which may result in long residence times for elements in the water in low-energy areas [20].

Plastics and contaminants may be of particular concern where they enter food webs at lower levels (e.g., plankton and forage fishes) and transfer contaminant loads to higher-trophic-level predators (e.g., killer whales, Orcinus orca) or commercially and culturally important species (e.g., Pacific salmon, Oncorhynchus spp.). The Pacific sand lance (Ammodytes personatus, hereafter sand lance), Pacific herring (Clupea pallasi, hereafter herring), and surf smelt (Hypomesus pretiosus, hereafter smelt) are important forage fish in this region [21,22]. Forage fish are protected resources, but their data are limited [23,24], particularly in the Salish Sea [25,26]. Nearshore habitats are critical to many forage fish for spawning, rearing, and foraging [27]. MPs may concentrate in fish nursery habitats nearshore and at spawning beaches. The sand lance may be uniquely susceptible to the impacts of MPs in marine habitats, since this species regularly occupies both benthic sediments and pelagic waters [28]. This species uses beach sediments, subtidal marine bedforms, and bottom substrates [29,30,31,32,33]. It buries itself on diel timeframes [28] and for extensive periods of time in winter [30]. Other studies have examined potential contaminants in sand lance in the Salish Sea, including persistent contaminants [34,35], hydrocarbons [36], and MPs [37,38,39].

We examined sediments at beach sites and subtidal bedforms known to provide critical habitats to sand lance and other forage species. Sampled areas included sediments in subtidal bedforms (known sand lance deep-water habitat) and exposed and protected shorelines (spawning habitat for sand lance, herring, and smelt) in the San Juan Archipelago (SJA) in the central Salish Sea and throughout Puget Sound in the southern Salish Sea. The analyses sought to determine if MPs were present; to document concentration, type, and size; and to compare regions and site characteristics. Complex physical processes determine where plastics accumulate. Understanding these processes may help to inform related phenomena, such as the size distribution of MP particles, prevalence of MP types, and relative densities of MPs in different areas. Our assumptions were that sites proximate to urban areas and nearshore sites would have increased microplastic loads relative to remote and offshore sites.

2. Material and Methods

2.1. Study Site

This study was conducted in the SJA in the central Salish Sea (Figure 1, Appendix A Figure A1) and contrasts MP concentrations on beaches in this area to those sampled in Puget Sound in the southern Salish Sea (Figure 1). Located at the confluence of the Strait of Georgia and the Strait of Juan de Fuca, the SJA is subject to large tidal currents and regular exchange with the open ocean [40]. In the SJA, we conducted sediment surveys at 3 sites confirmed as habitats for sand lance [41,42]. These included two beaches and an offshore benthic sand wavefield. Both beaches were also confirmed spawning sites for surf smelt and sand lance. The beach sites included a protected beach (Jackson Beach) within the San Juan Channel and an exposed beach (South Beach) on a peninsula that extends into the Strait of Juan de Fuca. Jackson Beach is located on the north side of Griffin Bay and is protected from winds, wave action, and currents; it also has local freshwater inputs. South Beach is at the southern tip of San Juan Island and is exposed to strong currents and tidal energy. We also investigated MP concentrations in subtidal sediments on the seafloor in the central San Juan Channel (Appendix A, Figure A2). In Puget Sound, we examined sediments at 12 beach sites. As the second largest estuary in the United States (25,000 miles of shoreline), its broad bathymetric range (0–280 m; https://pugetsoundestuarium.org/, accessed on 1 January 2025) and tidal fluctuations (delta tidal height 1–5 m https://tidesandcurrents.noaa.gov/, accessed on 1 January 2025) influence basin-specific residence times and net flow. Puget Sound was therefore compartmentalized into four basins—North Basin, Main Basin, Hood Canal and Admiralty Inlet, and South Sound—based on the sills in the region [43].

2.2. Sediment Sampling

All sampling occurred in a 4-month period in late fall and early winter (November–February) in the years 2016–2018. In the SJA, a 0.25 m² PVC quadrat (Figure 2) was placed just below the high tide or slack line at each beach site. The quadrat was then moved one meter down the high tide line and sampled again for a total of 10 samples per site. Samples were taken in low tide conditions. Within the quadrat, half of the surface area (0.125 m²) was collected for an individual sample. All materials to 3 cm depth were collected. The other half of the sample space within the quadrat was also collected in the same manner as a duplicate. The samples were stored in sediment sample bags. In Puget Sound, the same protocol was applied at 12 sites, with 10 samples per site at low tide, using a 1 m² PVC quadrat, applied just below the high tide or slack line at low tide. In the San Juan Channel subtidal sand wavefield, 10 samples were taken at an 80 m depth using a Van Veen grab sampler deployed from the R/V Centennial [30,42]. Beach and subtidal sediments in the SJA were analyzed for grain size metrics. Sediments were dried and processed in a Ro-tap (Combustion Engineering Inc., New York, NY, USA) sediment particle vertical sieve array to separate dried sediment samples into distinct component particle sizes [44]. Data were analyzed using GRADISTAT software, version 8.0 [45]. Mean, mode, sorting, skewness, and other statistics were calculated arithmetically, geometrically (in metric units), and logarithmically (in phi units) using the moment and Folk [46] graphical methods.

2.3. Sediment Processing for Microplastics

Analysis of all samples was conducted at the Limnology and Oceanography Lab at the University of Washington in Tacoma. The protocol and methods followed those published by the NOAA Marine Debris Program [47,48] (Appendix A, Figure A3). This method is appropriate for analysis of hard plastics, soft plastics, fragments, films, fibers, foams, and flakes. Samples were prepared by weighing a clean, dry 800 mL beaker; 400 g of wet sediment was weighed (precision to 0.1 mg), added to the beaker, and placed in a drying oven at 90 °C overnight. Once dry, the beaker and sediment were weighed to determine the dry sample weight. Density separation of the sample was accomplished by adding 300 mL of a 1.6 g/mL aqueous lithium metatungstate solution to the dried sediment in the 800 mL beaker. Disaggregation was achieved via stirring for 10 min using a magnetic stirrer with a glass surface at 120 rpm to float out any MPs. The sample was then poured through 5 mm and 0.3 mm sieves. Solids collected on the 0.3 mm sieve were transferred into a tared 500 mL beaker. This process was repeated until all floating debris was collected. Wet peroxide oxidation (WPO) was then performed in a fume hood, adding 20 mL of aqueous 0.05 M Fe(II) solution and 20 mL of 30% hydrogen peroxide to the 500 mL beaker. The mixture was left to stand at room temperature for 5 min before a stir bar was added and heated to 75 °C for 30 min. The beaker was removed and allowed to cool. This process was repeated until no natural organic material was visible. NaCl (6 g per 20 mL) was added to the 500 mL beaker to increase the density of the aqueous solution to 5 M NaCl and the mixture was heated to 75 °C until the salt dissolved. Another density separation was performed by transferring the WPO solution from the previous step to the density separator. The WPO beaker was rinsed with distilled water to transfer all remaining solids to the density separator, where it was loosely covered with aluminum foil and left overnight to allow the solids to settle. The next day, all floating solids were collected on a clean 0.3 mm sieve and allowed to air dry while loosely covered with aluminum foil overnight. To prevent contamination of the samples by airborne MPs during extraction, the materials were rinsed with filtered deionized water after every use and white cotton clothing was worn during all stages of the analysis.

2.4. Microscope Examination and Classification and Gravimetric Analysis

Visual sorting is one of the most commonly used methods for the identification of MPs, including type, shape, and color as qualitative categories [47]. The MPs were examined using a dissecting microscope at 40 X magnification and visually identified, examined, quantified, and categorized. The MPs were classified and reported by color and type (fragments, fibers, films, foam, flakes) [49,50,51] (Figure 2). Measurements were conducted on the largest dimension. The number of each type of MP and total individual pieces collected were recorded for each sample in both MPs per gram and MPs per meter² to facilitate comparisons to results in other studies. Material type was confirmed via melting point [52,53,54]. A 60 w pencil-type temperature-adjustable (200–450 °C) soldering iron (1.0 mm extra-fine tip) set to 350 °C was used to assess material type through direct contact (1–2 s). Materials were processed individually, placed on a glass slide, and probed with the soldering iron, all while observed through a microscope. MP responses to heat included softening, melting, and/or deformation, in contrast to cellulosic fibers, which readily ignited, and items of metallic or inorganic origin, which had no response to the application of heat.

2.5. Classification of Shorelines and Oceanographic Attributes

Shoreline habitat classifications, sediment type, and wave exposure were determined according to a detailed inventory of linear shoreline data updated in 2014 [55] (Washington State ShoreZone Inventory; https://www.eopugetsound.org/maps/washington-state-shoreline-habitat-classes, accessed on 1 January 2025). Site status as a spawning location for smelt, herring, and sand lance was based on Washington Department of Fish and Wildlife data (WDFW Habitat Program, forage fish spawning map; https://wdfw.maps.arcgis.com/apps/mapviewer/index.html, accessed on 1 January 2025). Current speed (max, mean, and standard deviation; cm/sec) and mean current direction were calculated from annual prediction tables of tidal currents, developed through the Operational Forecast Systems developed at the National Oceanic and Atmospheric Administration (NOAA, National Ocean Service, Center for Operational Oceanographic Products and Services, https://tidesandcurrents.noaa.gov, accessed on 1 January 2025). Tidal current values were estimated as the absolute value of thousands of ebb and flood values predicted throughout the year. Linear distance over water from urban centers was calculated using GPS tools in Navionics (https://www.navionics.com/usa/, accessed on 1 January 2025). Urban population size was determined via census data in Statistics Canada and the US Census Bureau reports (Victoria = 92,141; Everett = 110,812; Seattle = 733,919; Tacoma = 219,205; Olympia = 55,731; retrieved April 2024).

2.6. Statistical Analysis and Models

The data analysis software R (version 4.5.1) was used to conduct Chi-squared and one-way Analysis of Variance (ANOVA) tests. Chi squared was used to determine statistical significance in the frequency of occurrence of plastic between sample locations. ANOVA and Tukey HSD post hoc tests were conducted to determine if plastic concentration varied between sites. Generalized additive models (GAMs) [56] were applied to evaluate the contribution of current, sediment type, wave exposure, and population size and distance from the nearest major urban area to the MP loading in beach sediments. GAMs were applied to determine the additive contribution and marginal impact of each variable and to allow for nonlinear relationships between the independent and dependent variables; the models were applied using a generalized cross validation (GCV) estimation process (mgsv package, 1.9-1) [57,58].

3. Results

3.1. Microplastics in Various Sites

MPs were found in every sample tested. MP loads differed between regions (F_1,12 = 4.20, p = 0.062), with MPs significantly elevated in Puget Sound compared to the SJA (Figure 3). In the SJA, the protected beach had a mean concentration of 16.77 MP/m² dry weight (DW), the exposed beach had a mean concentration of 12.33 MP/m², and the subtidal wavefield had a mean concentration of 9.01 MP/m² (

Table 1. Microplastics in sediments.

Site	MP/kg DW	MP/m2	Source
Research Sites—Salish Sea, Pacific Ocean
San Juan Archipelago Protected Beach	134.8	16.77	**
San Juan Archipelago Exposed Beach	45.4	12.33	**
San Juan Channel sand wavefield	43.6	9.01	**
Puget Sound, Howarth Beach		312	**
Puget Sound, Carkeek Park		285	**
Puget Sound, Alki Beach		249	**
Puget Sound, Dash Point State Park		138	**
Puget Sound, Discovery Park		135	**
Puget Sound, Golden Gardens Park		135	**
Puget Sound, Elliot Bay Marina		126	**
Puget Sound, Picnic Point		114	**
Puget Sound, Tolmie State Park		111	**
Puget Sound, Shine Tidelands State Park		102	**
Puget Sound, Edmonds Marina		39	**
Puget Sound, Mukilteo		30	**
Reference Sites—Northeastern Pacific Ocean
Puget Sound, Salish Sea, Pacific Ocean	150.0		[59]
Puget Sound, Salish Sea, Pacific Ocean	60.2		[60]
Coastal British Columbia, Pacific Ocean	max = 9200		[9]
Strait of Georgia, Salish Sea Pacific Ocean	30–230		[61]
Reference Sites—Northeastern Atlantic Ocean
Baltic Sea, Atlantic Ocean	1.3−36	370−7330	[62]
Baltic Sea, Atlantic Ocean	3235		[63]
Baltic Sea, Atlantic Ocean	2–572		[64]
Baltic Sea, Atlantic Ocean	100–1000		[65]
Baltic Sea, Atlantic Ocean	34–10,179		[66,67]
Reference Sites—Global Oceans
Worldwide, Multiple Oceans	115	3155	[68]
Worldwide, Multiple Oceans	1.3–10,179	1–336,000	[69]
Worldwide, Multiple Oceans	1–2000		[70,71]

Notes: Results from this study are indicated (**, first 3 rows), where microplastics in sediments are estimated per kilogram dry weight and surface density of distributed microplastics (m²). Values from other analyses of microplastics in marine sediments elsewhere are provided for comparison.

). Significant differences were noted in the number of plastics per sample at each site (ANOVA, F_2,8 = 6.99, p = 0.018), where the protected beach had an elevated abundance of MPs (mean = 27.25 ± 12.31 SD) relative to both the exposed beach (mean = 9.25 ± 3.86 SD) and the subtidal sand wavefield (mean = 9.00 ± 1.58 SD). Pair-wise comparisons of differences between sites detected significant differences between the protected beach and both the exposed beach and subtidal sand wavefield (Tukey HSD, p < 0.031); no differences were noted between the exposed beach and subtidal sand wavefield (Tukey HSD, p = 0.978). In Puget Sound, MP concentrations were significantly higher (mean = 148 ± 89.1 SD, range =30–312 MP/m²), with the lowest concentration approximately ×2 and the highest concentration nearly ×20 higher than the highest-concentration site in the SJA. Although the results indicate differences between the individual sample sites in Puget Sound, most sites sampled in Puget Sound were in the Main Basin (N = 8) and significant differences were not evident in comparisons to the sites in the other basins (South Sound N = 1; Hood Canal/Admiralty Inlet N = 1; North Basin N = 2; ANOVA).

3.2. Types of Microplastics

Plastics were classified according to their type—fragments, fibers, films, foam, and flakes. At the sites in both regions, fibers were the most abundant MP found. Additionally, fibers were found in almost every sample analyzed. In SJA sites, 87% of MPs were fibers, 8% were films, and 5% were fragments. Relative rates of the three types of MPs were similar between the sites (χ² _{df = 4} = 1.40, p = 0.844; Figure 4). At the protected beach, 87.2% of MPs were fibers, 7.3% of plastics were films, and 5.5% were fragments. At the exposed beach, 86.5% of MPs were fibers, 10.8% were films, and 2.7% were fragments. The wavefield had 77.8% fibers and an equal proportion of films and fragments (11.1%). In Puget Sound sites, MP deposits were also predominantly fibers, comprising approximately 73% of the MPs collected (Figure 4). Films were 2% of the total. Flakes were the predominant type at Howarth Park in Everett (proximate to a naval shipyard) and were also present at two other locations—Carkeek Park in Ballard (commercial shipyard) and Elliot Bay (marina and port for cruise ships); the flakes were identified as an epoxy-based paint used for sealing ship hulls.

3.3. Size Distribution of Microplastics

The size of the MPs identified ranged <1 mm–10 mm. The vast majority of MPs (89%) were between <1 mm and 5 mm. To investigate potential patterns in size-related depositions for beach sites and subtidal sites, the relative size of MPs was also examined at the SJA sites. Relative rates of the size classes of MPs differed between sites in the SJA (Figure 5). In contrast, the relative ratio of size classes was similar between the protected and exposed beach (χ² _{df = 3} = 1.66, p = 0.646), and the wavefield differed substantially (χ² _{df = 3} = 8.72, p = 0.035). Notably, at the subtidal benthic wavefield, there was a higher relative proportion of MPs (MPs < 1 mm), and no plastics in the largest size class (5–10 mm) were found at this site. At the protected beach, 38% of MPs (MP density = 51.5 MP/kg DW) were in the smallest size category (<1 mm), 26% (34.5 MP/kg DW) were 1–2.5 mm, 23% (32.0 MP/kg DW) were 2.5–5 mm, and 13% (16.8 MP/kg DW) were in the size range 5–10 mm. At the exposed beach, 46% of MPs (21.0 MP/kg DW) were in the smallest size category (<1 mm), 18% (8.2 MP/kg DW) were 1–2.5 mm, 27% (12.4 MP/kg DW) were 2.5–5 mm, and 9% (3.8 MPs/kg DW) were in the size range 5–10 mm. At the subtidal wavefield, 67% of MPs (29.0 MP/kg DW) were in the smallest size category (<1 mm), 10% (4.6 MPs/kg) were 1–2.5 mm, 23% (10.0 MP/kg DW) were 2.5–5 mm, and none were in the size range 5–10 mm (Figure 5).

3.4. Analyses of Site-Specific Wave Exposure and Sediment Type

The data available at the SJA sites included quantitative metrics for sediment type and grain size distribution. These data were used to analyze differences in grain size distribution and other sediment metrics between the beach sites and the subtidal site (Figure 6). Both the exposed beach site (South Beach) and the subtidal site (San Juan Channel) were composed of well-sorted sediments, unimodal in sediment particle size distribution, and comprising coarse sand and gravel. The protected beach site (Jackson Beach) was composed of poorly sorted sediments and was polymodal in its sediment particle size distribution. The data available at the Puget Sound sites enabled analyses of site-specific differences in wave exposure, embayment type, and sediment type, using categorical or qualitative classifications of each site in wave exposure. In analyses of MP concentrations in the Puget Sound sites, neither wave exposure (ANOVA, F_1,10 = 1.61, p = 0.146) nor embayment type (ANOVA, F_1,10 = 1.52, p = 0.245) were significant factors, though lagoon sites were elevated (mean MP/m² = 219) compared to partial enclosures (mean MP/m² = 134). Sediment type in the Puget Sound sites was not significant (ANOVA, F_3,10 = 2.37, p = 0.147), and no significant differences were noted between sediment classifications (Tukey HSD, p > 0.118), though the results suggest that artificial sediments had the highest MP loading (mean MP/m² = 312), followed by sand (mean MP/m² = 155.1), mixed fine sediments (118.5), and mixed coarse sediments (mean MP/m² = 70.5; Figure 7).

3.5. Microplastics at Beach Sites Confirmed as Spawning Habitats for Forage Fish

Most sites in the SJA and in Puget Sound are documented as spawning sites for sand lance (N = 10), surf smelt (N = 9), and/or herring (N = 4). Sites confirmed as spawning habitats for forage fish were high in MPs (Figure 7). In Puget Sound, all of the top five sites with the highest MP loads were sites documented as spawning beaches for sand lance and surf smelt, and the four sites that supported herring were among these top five sites (>135 m²). In all three species, MP concentrations were elevated in spawning habitats (mean MP/m²; smelt = 184.88; herring = 201; sand lance = 171.75) relative to non-spawning habitats (mean MP/m²; smelt = 74.25; herring = 121.5; sand lance = 100.5).

3.6. Model of Microplastic Loads as a Function of Site Attributes (Oceanography and Sediments)

A GAM was applied to all data, using quantitative metrics available related to potential sources and distribution mechanisms for MPs, including (1) variables for current speed (max, mean, SD), (2) urban population size for the closest urban area, (3) distance from the closest urban area, and (4) distance from the closest shipping lane. Models were fitted using a Poisson distribution (link = log). The best-performing model included the mean current speed and urban population size for the closest urban area (adjusted R² = 0.351, Deviance explained = 52.2%).

4. Discussion

MPs are widely dispersed in the marine environment and are present throughout the water column, on beaches, and on the seabed [10,11]. MPs are also prevalent in important fish habitats and in marine organisms and present a potential threat to the health of marine ecosystems [72].

4.1. Relative Abundance of Microplastics

In our analyses, MPs were found in every sample analyzed and at each location. Even samples taken of sediments 80 m below the surface in a high-flow environment [17] contained MPs. This benthic feature is a large bedform of unconsolidated sediments—coarse-grain sand and shell hash fragments [15]. In our comparison of beach sites, the exposed beach site had the lowest overall concentration of MPs. This area is exposed to high tidal exchanges, strong wave energy, and alongshore hydrodynamics [73], all of which may prevent deposition and remove MPs that might otherwise accumulate over time.

4.2. Microplastic Types and Size Classification

All three types of MPs classified in this research—fibers, films, and fragments—were present at each sample site. Overall, the vast majority were microfibers. This finding further confirms previous results that suggest microfibers are the most abundant MP in the Salish Sea [1,9,61]. Black et al. [59] showed that the majority of MPs collected in shallow water near storm water outlets in Puget Sound were fibers, and Harris [74] reported that buoyant fiber-shaped particles comprise approximately 90% of MPs on beaches. Many studies elsewhere have also identified the vast majority of MPs to be fibers [5,75,76]. The relative proportion of films, fibers, and particles was also similar to that reported in Esiukova et al. [64].

While both beaches were similar in their relative proportion of MP types, the subtidal sand wavefield had a higher proportion of fragments. This may be because the high flow and constant tidal exchange in this area reflect processes that not only retain sand particles of a certain size and shape but also contribute to the accumulation and/or retention of MP particles that are similar in size to the sediments within the wavefield. Singdahl-Larsen [77] found that while fibers were the dominant form of MPs in the upper sediment layers, other forms (films and fragments) dominated in the lower sections of sediment cores. It may be that the constant shifting of MPs in the subtidal sand wavefield exposes more of these particles that might otherwise be present only in lower sections of accessible sediments.

Wide size ranges of MPs were found. MPs ranged between <1 and 10 mm, with 89% of MPs ranging between <1 and 5 mm. No larger MPs were found (MPs were <1–5 mm). We attribute this finding to larger pieces accumulating on beaches while only smaller MPs descend through the water column to the sea floor. The most common size of MP found overlaps the size of zooplankton found in the sand lance diet; the most abundant size of the MPs was <1 mm–5 mm, while calanoid copepods (Calanoida spp.) range ~0.7 mm–1.2 mm [43].

4.3. Comparison to Microplastics in Sediments Elsewhere in the Salish Sea

Coastal ecosystems are thought to be particularly vulnerable to MPs because they enter the marine environment from terrestrial pollution via urban outflow sites and river inputs [1,78]. Several urban areas are in close proximity to our study sites, including Vancouver, Victoria, Bellingham, Everett, Seattle Tacoma, and Olympia. There are also several major rivers close to the beaches in this study, including the Fraser, Squamish, Skagit, Stillaguamish, Snohomish, Duwamish, Puyallup, Nisqually, and Deschutes rivers. In the SJA, there is high flow as the area is surrounded by straits with significant tidal exchange [17]. The sediment samples examined in the SJA were relatively low in MPs compared to the more urbanized and more enclosed Puget Sound basin (mean prevalence of 1068 MP/m² in sediment on Seattle beaches and 606 MP/m² in sediment on beaches outside of Seattle city limits), in contrast to mean concentrations of 13 MP/m² in our analyses of the beaches in the San Juan Islands. The Main Basin in Puget Sound has the highest population density due to the two major cities on its east side, Seattle and Tacoma. King County (encompassing Seattle) also has a population of 2,188,649 compared to San Juan County (SJA), which has a population of 16,715 [79]. The Seattle area is also characterized by relatively retentive inland waters. The presence of relatively high concentrations in Puget Sound parallels studies by Black et al. [59], who found high concentrations of MP fibers in nearshore sediments (mean of 150 fibers/kg DW). Collicutt et al. [60] conducted MP analyses of both water and sediment samples, measuring 659.9 ± 520.9 MP/m³ and 60.2 ± 63.4 MP/kg DW, respectively. In coastal BC, the average MP concentration in surface waters has been estimated at 0.59 MP/m³, and while no clear spatial pattern was evident in the analyses conducted by Mahara et al. [80], Desforges et al. [9] found MP concentrations in the western coast of British Columbia to be highest in nearshore areas (up to 9200 MP/m³).

4.4. Comparison to Other Analyses of Sediment Loads in Beaches and Subtidal Bedforms

A review of 32 studies from 2015–2020 in the Baltic Sea, Black Sea, Barents Sea, Kara Sea, Bering Sea, Sea of Okhotsk, and Sea of Japan noted that MP prevalence levels vary widely: from 0.6 to 336,000 MP/m³ for MPs in water and from 1.3 to 10,179 MP/kg (DW) in sediments [69]. Our results suggest that MP concentrations are likely to be higher at shallower depths and in sediments with low proportions of very fine sand sediment. This finding has also been observed in other sedimentary environments [66]. One possible reason is that MPs accumulate in coarse, sandy sediments because they enter the environment in sizes comparable to larger sediment grain sizes.

MPs in marine sediments and coastlines are highly prevalent yet also variable in time and space [81,82]. Variation has been shown to be especially high for large marine litter on the coastline [83]. Plastic litter can be found both close to population centers and on remote shorelines [10,84], though relatively large marine litter is typically more abundant near coastal cities [83,85], fishing ports and navigation routes, land-based point sources, rivers and canal discharges, and public beaches [85,86]. Variability is caused by not only the proximity of sources of pollution but also by the patchiness due to local hydrodynamics and other physical processes [87,88]. Concentrations of MPs at beaches range from 1038 to 7070 items/m², and on heavily polluted beaches, MPs (0.25–10 mm) can make up 3.3% of the sediment by weight [68]. The global mean MP (0.5–5 mm) concentration amounts to 3155 ± 1308 MP/m² or 115 ± 61 MP/kg DW [68]. Esiukova [62] reported MP contamination in beach sediments at a gross mean value of 1.3−36.3 MP/kg (or 0.05−2.89 MP/g of dry sediment or 370−7330 MP/m²); other European studies report MP concentrations from 1 MP/kg DW to over 2000 MP/kg DW [69,89]. Chubarenko et al. [63] report a median content (MP/m²) of 0.85 for macro-, 1.48 for meso-, and 3.35 for large microlitter, and 3235 for MPs (0.5–5 mm). Eo et al. [85] also discriminate by size class and report a statistically distinct spatial distribution of plastics according to size on beaches. Chubarenko et al. [63] similarly distinguish the distribution of litter and MPs along and across beaches.

Fewer studies have examined subtidal marine sediments. Studies on MP concentrations in marine bottom sediments report hundreds of items per kilogram of dry weight (range of mean values: 34 ± 10 MP/kg DW to 10,179 MP/kg DW) [64,66,67] and 100–1000 MP/kg DW [65]. More research is needed to determine oceanographic and biological processes that might inhibit or increase vertical transport to the seabed [13,90].

4.5. Sources and Transport of Microplastics

4.5.1. Sources

MP pollution is brought to the ocean in many ways, and linking MPs to a specific source is challenging. MP densities vary greatly by location. Plastics may enter the marine environment via terrestrial sources (e.g., rivers, runoff, wastewater), via marine sources (e.g., boats, fishing gear), or via atmospheric deposition [91]. MP pollution in sediments has been linked to proximate human activities [83], with high concentrations of MPs in the sediments of coastal harbors [92,93]. Landfills, recycling and industrial facilities, and municipal wastewater are other significant sources of MPs [94]. In our analyses, MPs were most abundant at Jackson Beach, which is relatively more protected from the turbid open water and strong wave energy of the San Juan Channel and from the straits north and south of the archipelago. These relatively calmer waters might allow the plastics to accumulate on the shore instead of dissipating in the channel; this assumption is also supported by data that show a wider variety of sizes are present at Jackson Beach. Offshore in the subtidal bedform, which is in the center of the channel and exposed to high flow and tidal exchange, no MPs larger than 5 mm were found.

4.5.2. Movement and Transport in Pelagic Environment

The mechanisms that drive the horizontal distribution of MPs include prevailing currents, geostrophic circulation, and wind [93]. More than half of produced plastics have a lower density than seawater, which facilitates passive floatation at the surface [3], and particle aggregation or biofouling influences vertical transport or deposition. Wind-induced turbulences such as breaking surface waves, bubble injection, and Langmuir circulation also influence vertical distribution [3]; intrinsic factors (e.g., density, size, and shape) also influence deposition through advection velocity and buoyancy relative to turbulent displacement [78]. The Salish Sea is a large, complex estuarine system that flushes on timescales of several months [18,40]. Circulation patterns in Puget Sound vary in each basin. The mean residence time for Puget Sound is estimated at 90 days (range = 20–120 days) [95]. The Main Basin has a mean residence time of 22 and 38 days for surface and deep water, respectively; the Hood Canal has a mean of 8 days for surface and 15 days for deep water; and South Puget Sound has a mean of 24 days for surface and 23 days for deep water. Although residence times in the Main Basin are relatively faster than in the South Sound Basin, there is low turbidity in the bottom water of both basins to stir up and horizontally transport the sediment and any imbedded MPs. In the SJA, flow is much stronger due to the confluence of vertical mixing of the water bodies moving from the Strait of Georgia to the Strait of Juan de Fuca and direct estuarine circulation with the Pacific Ocean. The primary control of the physical oceanography of the Georgia–Fuca System (Strait of Georgia, the San Juan Islands, and the Strait of Juan de Fuca) is estuarine circulation. Forced by the pressure gradient, the net flow moves out of the estuary to the ocean. This region is subject to strong mixing events and tidal currents (1–2 m/s), leading to a near-homogenous vertical structure. Flows are influenced by frictional effects and nonlinear interactions. Intense mixing is evident at the subtidal site in the San Juan Channel, with high velocity shear related to high exchange funneled through narrow topography. High wave action and mechanical dissipation are also evident at the exposed beach site in the SJA.

4.5.3. Patterns in Deposition and Relative Abundance in Marine Habitats

A global analysis of marine litter data showed that the proportion of plastic increased progressively from 49% on riverbeds to 64% on nearshore bottoms (<100 m depth, <100 km from shoreline) and 77% on deep seafloors (>100 m depth, >100 km from shoreline) [96]. This suggests that plastics have greater persistence than other types of marine litter. In high-energy coastal and shelf environments, macroplastic items rapidly break down into smaller MP particles via mechanical fracturing during bedload transport. MPs and less dense particles settle to the seafloor in low-energy depositional environments (e.g., fjords, lagoons, estuaries) [71]. Harris [74] reported the highest median concentration of MP particles in fjords at 7000 particles kg−1 dry sediment, followed by 300 in estuarine environments, and 200 particles kg−1 in beaches and in shallow coastal environments. The environments that appear to contain the lowest numbers of MP particles are high-energy, non-depositional sedimentary environments, such as rocky shores, high-energy (storm and tide-dominated) continental shelves with coarse sand and gravel, and tide-dominated estuaries that tend to export fine-grained sediment. This is in accordance with our findings that coarse-grained, higher-energy sediments (e.g., San Juan Channel sand wavefield) do not retain MPs at high densities. Physical oceanographic drivers (e.g., currents, tides) control MP distribution within Puget Sound [95,96,97]. Coastal environments, in contrast, particularly low-energy sites, exhibit higher MP concentrations.

4.5.4. Deposition and Movement in Beaches

Beaches have the potential to be a good representation of MP accumulation over time because beaches are the interface between coastal waters and land-based sources [98]. Further work is needed to determine how MPs accumulate and spread on beaches. Accumulation of MPs on beaches will differ depending on whether the beach is a net sediment sink (i.e., depositional sedimentary environment), net erosional (retreating coastline with sediment loss), or in equilibrium in terms of sediment supply and removal processes [74]. Comparative studies might also be conducted to determine the accumulation dynamics of MPs along gradients of wave exposure and tidal height. Given that beaches and subtidal coastal habitats are dynamic, with continuous and seasonal erosion of sediment, MPs may become buried in sediment during periods of accretion. Furthermore, beaches filter and retain particulate organic matter over a range of depths, and sediments between a 0 and 5 cm depth are characterized by steep gradients and strong seasonal variations in more fine-grained particles and particulate organic matter [47]. Dense populations and increasing urbanization of cities located near either coastlines or other marine bodies of water may result in greater amounts of anthropogenic pollution entering the environment, and differences in plastic concentrations between sample sites are likely due in part to anthropogenic factors.

4.5.5. Deposition and Movement in Subtidal Sediments

Studies in subtidal areas have revealed that MPs are more abundant in subtidal sediments than on sandy beaches and in estuarine habitats [83,89,99]. While this differs from our findings, our results reflect correlations between plastic particle size and wave and current energy reported elsewhere [100,101]. The fate of plastics might be expected to be similar to that of naturally occurring organic particles and mineral particles (e.g., sand, silt) and governed by the same physical laws, with hydraulically equivalent physical properties [101]. Plastic particles with different densities are mobilized at the same level of bed shear stress (i.e., hydraulic equivalence) needed to mobilize smaller-sized quartz grains [74,101]. Alternatively, due to their low densities, most plastic particles <2 mm in size might be expected to be transported as suspended loads rather than bedloads. Better understanding these properties would be important in the context of our research and highly relevant to predicting MP loads in sand lance and sand eel habitats. Typically, these habitats are in high-velocity environments that winnow sediments and retain coarse grain (e.g., sand, gravel), while removing fine-grain sediments (e.g., silt and clay). These attributes describe the site we examined [32,41], where, due to tidal wave deformation (enhanced asymmetry in ebb and flood tidal currents), relic glacial sediments were highly sorted. These processes are relevant in other macrotidal ebb and flood-dominated channel systems in the region important to sand lance. In mud and silt bedforms, however, MPs in sediments may be retained longer due to their ability to form aggregates with organic matter [102], which may cause a faster rate of sinking from the water column, higher accumulation on the ocean floor, and retention in low energy habitats. Further research might examine differences in MP loading across mud, sand, and rock habitats and implications for marine species associated with these habitats.

4.6. Microplastics in Critical Fish Habitats

Not all habitats are equally important to marine species. It is perhaps most important to examine MP concentrations in habitats known to be critical to marine life. The near ubiquity of MPs in coastal sediments [100] and their importance to marine life in many forms clearly designates these areas as important for assessment and monitoring. Further analyses of subtidal and benthic habitats known to support marine life are also critical. In the context of sand lance, many habitats have been mapped. Habitats known to be used consistently and repeatedly [32,103] should be monitored more frequently to determine how MP loads might shift in finite (e.g., tidal) as well as extended (e.g., annual, decadal) timeframes. Elsewhere in the Salish Sea, Peters [61] detected significantly higher concentrations of MPs in habitats deemed suitable to sand lance (mean = 460 MP/kg; median = 230 MPs/kg) than habitat deemed unsuitable (mean = 120 MP/kg; median = 30 MP/kg). MPs were also found in higher concentrations in shallow (<40 m) rather than deeper areas (>40 m), with increased MP loads in sediments with very fine sand (grain sizes between 0.125 mm and 0.063 mm) [61]. Further research might more definitively categorize benthic habitats and groundtruth sand lance presence or absence [104]. Further research might also explore how oceanographic processes influence MP deposition and distribution as well as how oceanographic processes influence species distribution and movement patterns [104]. Cross-mapping the distribution of plastics and other benthic and demersal associated species, especially commercial fish stocks, might provide a useful evaluation of exposure and risk of ingestion.

4.7. Implications for Marine Life

MPs transfer to marine organisms through absorption or ingestion and have the potential to transfer through the food web [105]. MPs themselves may be harmful as a mechanical hazard [90,106], and the transfer of micro-organisms or associated pollutants that affect physiological functions [99,107] may increase stress in fish and marine life [108] and influence their growth, development, sexual maturation, and reproductive capacity [109]. More research is needed to assess the population status of data-poor species and habitats [23] and to identify potential threats [25]. Life stage may also be important in determining impacts. Foraging dynamics; growth and condition shift across seasons [30] and throughout life stages [110]; and patterns in MP accumulation (as well as flushing and elimination) may reflect these dynamics. Since sand lance undergo sexual maturation while dormant in benthic sediments [111], exposure may increase due to MP loads in sediments.

Sediment-dwelling organisms are sensitive indicator species and are often used as bio-indicators of ecosystem health [112,113], including benthic and demersal fish [114,115]. Monitoring fish diets provides insight into species interactions and predator–prey dynamics [116]. In many cases, coordinated sampling already exists to monitor diet in important commercial fish stocks [117]. Where absent, there may be opportunities to collaborate with recreational or commercial operators [118,119] to collect data and stomach contents and inform analyses of trophic transfer and contamination rates at various trophic levels, life histories, and locations [39].

4.8. Methodological Approach and Limitations

Sampling, extraction, and detection methods and analytic techniques are constantly being developed and further refined worldwide [120]. Increasingly, new protocols and procedures are available to ensure quality control and determine microplastic origin and composition [121]. Our analyses did not include polymer identification (e.g., via FTIR or Raman spectroscopy), and distinguishing plastic and organic compounds via melting point may result in false positives. Additionally, no procedural blanks or field blanks were used. Subsequent to the processing of these samples, several studies have recommended procedural blanks be used to monitor potential sample contamination [121,122,123]. These approaches were not used in this study, but have been used in subsequent studies in the same facilities in similar analyses and have shown 0% external contamination in processing (M.R. Baker, unpublished data). There were also limitations on identification. Visual identification at 40 × magnification is insufficient for particles <300 µm and is therefore susceptible to bias. Currently, a wide range of sampling techniques are used for monitoring MPs in sediments [47,124], including areal surface sampling using a quadrant (abundances per unit of surface) [125] and three-dimensional samples using vertical pits [63,64] (abundances in volumetric metrics [78,99] or according to weight [89]). MPs in bottom sediments have been examined using box corers [126] and van Veen grab samplers [63,65,66,67]. A few additional limitations must be noted. In the case of the van Veen sampler, vertical structure of sediments was not conserved and analyses of the vertical distribution of plastics were not possible. On beaches, sediment samples might also be extracted from different locations or zones (e.g., high-littoral, sub-littoral zone). Chubarenko et al. [63] and Esiukova et al. [64] report a wide range of MP densities from wracklines to the waterline (range = 2 to 572 MP/kg DW; mean = 108 MP/kg DW). Additionally, our analyses did not consider other potentially relevant factors, such as sediment organic content, seasonal variation, biological activity (e.g., bioturbation), or factors that might influence biodegradation [109]. Future work may investigate contamination as a function of beach zone and depth distribution as well as seasonal variation and the aforementioned processes. Additionally, the asymmetry of the samples between sites (SJA, N = 3; Puget Sound, N = 12) limits the strength of the comparative analysis; future sampling should increase sampling in the SJA area.

4.9. Suggestions for Future Research

MPs are present in the marine environment worldwide [99,127,128] and sediments are hypothesized to be important sinks for these pollutants [84]. In determining the fate and threat of MPs in the marine environment, more attention should be directed to how contaminants accumulate in habitats. Further defining benthic habitats is critical [129]. Physical processes (e.g., currents, tides, flow regimes), bathymetry, and benthic substrates may also be further explored in the context of where MPs accumulate. Research should also investigate potential biological transport [130] in addition to delineating and modeling processes related to deposition, degradation, and physical transport.

5. Conclusions

Our analyses provide baseline data on MP presence and abundance in marine sediments and beach sediments in the ecologically important and oceanographically complex Salish Sea. MPs were found in all samples and sites and were higher in Puget Sound compared to the SJA. Microfibers were the most prevalent, though flakes were common proximate to shipyards and marinas. Protected beaches were elevated relative to exposed beaches and subtidal sediments and MPs were elevated in sand and fine grain sediments, poorly sorted sediments, and artificial sediments. The model results indicate that current speed and proximate urban populations influence nearshore microplastic concentrations. The beaches in this region are important to forage fish, and MPs were elevated at sites confirmed as spawning habitats. Our research provides new insights into MP distribution, deposition, and retention in marine sediments and potential exposure in shoreline habitats. Further analyses are required to examine the relative influence of urban populations and shipping lanes, and how physical processes such as wave exposure, flow regime, tides, currents, beach aspects, and shoreline geometry influence the accumulation and retention of MPs. Further research might also investigate MPs in offshore banks, mud flats, and sediment fields throughout the Salish Sea to determine how MPs are transported, retained, and accumulated in substrates, as well as their potential impacts on marine biota.