Presence of microplastics and microparticles in Oregon Black Rockfish sampled near marine reserve areas

Subadult rockfish collection

The subadult rockfish were sampled in nearshore waters off the Oregon Coast from two distinct sampling efforts: samples collected in cooperation with a local recreational charter fishing company in the port town of Newport, Oregon (Charter Fish) and samples collected by the Oregon Department of Fish and Wildlife (ODFW Fish) in proximity to four geographically distinct marine reserves as part of hook-and-line monitoring activities (Fig. 1). Oregon Marine Reserves are fully protected areas meaning that ocean development and fishing are not allowed (ODFW, 2022). Charter samples were dissected in the field after being fileted, while the ODFW fish samples remained whole and frozen in plastic bags until dissection. Once in the lab, the ODFW fish were wiped with ethanol prior to dissection. Since we could not wipe the Charter samples with ethanol, we laid out a background contamination filter while dissecting the fish to address possible contamination in the field. The Charter samples were collected in summer 2018 and spring of 2019, while the ODFW samples were collected in the spring and fall of 2018 and 2019.

Overall, we processed a total of 58 subadult samples (GI tracts dissected from the whole fish) from across these areas near marine reserves (Cascade Head—10, Cape Perpetua—13, Cape Falcon—10, Redfish Rocks—11) and Newport, Oregon. GPS coordinates were noted for each fish collected by ODFW, showing the distance the fish were collected from a specific marine reserve. The average distance the fish were collected from their respective marine reserve ranged approx. between 6.8 km and 16.7 km. Although many Black Rockfish tend to have a high degree of site fidelity (Green & Starr, 2011), they have also been shown to move among sites (Parker et al., 2007), and therefore, we consider these sample fish to generally represent populations found in Oregon. ODFW does not need a collection permit for collections deemed relevant for research and management purposes.

Pelagic juvenile rockfish collection

Juvenile samples were collected through a collaboration by Oregon State University, the Oregon Department of Fish and Wildlife’s Marine Reserves Program, and the Oregon Coast Aquarium at two sites along the central Oregon Coast: Otter Rock, a marine reserve, and Cape Foulweather, a non-marine reserve site that is similar to the marine reserve in habitat, depths, oceanography, and historical fishing pressure (Fig. 1). While similar in habitat and topography, there is a difference in activities permitted between the reserve and comparison area. A comparison area is a separate location from the marine reserve, but in the vicinity of the reserve. These are areas where fishing and ocean development are allowed and also where monitoring occurs, while reserves prohibit fishing and ocean development. (ODFW, 2022). The fish were captured using Standardized Monitoring Units for the Recruitment of Fishes (SMURFs). Each SMURF consisted of low-density polyethylene (Fig. S2) mesh folded into a cylinder using garden fencing to form a structure similar to a kelp canopy, thus simulating natural structures which juveniles are known to settle to (Ottmann et al., 2018). The samples were retrieved by snorkelers using a butterfly net to enclose the trap while it was brought on board the vessel. Juvenile fishes sorted from the SMURF were immediately euthanized using 2 mM tricaine methanesulfonate (MS-222) buffered in 6 mM sodium peroxide. The samples were identified to species and frozen for later laboratory use (Ottmann et al., 2018). Overall, 66 juvenile Black Rockfish samples were processed for microparticles by inspection. Juvenile samples were collected under National Marine Fisheries Service (NMFS) permit #18058, and juvenile fish protocols were approved by Oregon State University Animal Care and Use Protocol #4183.

Dissection and microplastic analysis

Dissection and microscopy procedures were conducted in a laminar flow hood (Erlab Captair Flow 391; Erlab USA, Rowley, MA, USA) (230 m³/h processed air flow) with a HEPA filter or a fume hood with HEPA filter, using a Leica EZ4W scope with an attached camera or a Leica EZ4 dissecting microscope and added camera (Moticam 3.0 camera) at 8-35X magnification, without the additional camera zoom (Leica, Letzar, Germany). A fume hood was utilized only briefly while working with hazardous chemicals. Otherwise, a laminar flow hood with HEPA filtration was used the majority of the time to prevent microplastic contamination.

All fish (sub-adult and juveniles) were weighed pre-dissection except for the Charter fish as they were filleted before we dissected them. Subadult fish were used if they have a total length of 34–45 cm. The average weight of the subadult fish from ODFW were 1,044.1 g. We dissected the subadults and extracted and measured (length and weight) the GI tract from the esophagus to the vent. Additionally, undigested contents in the stomach were removed. Due to the juveniles’ reduced size, all organs collected from the esophagus to the vent were digested for microplastic examination. We placed the GI tract of subadult fish into mason jars with the interior lining of the lids facing outwards, while juvenile guts were preserved in 20 ml glass vials. For subadult fish, we added 20% potassium hydroxide (KOH) at approximately three times the wet weight of the GI tract or 100 ml minimum. We digested juvenile samples in approximately 5 mL of the 20% KOH solution, which was roughly equivalent to the relative volume used for subadults. All samples were digested for 48–72 h in a water bath at 50 °C (either a Fisher Scientific Model Isotemp 220, Waltham, MA, USA or a Precision Scientific 180 Series Water Bath, Chicago, IL, USA), although some of the subadult samples were able to continue digesting at room temperature after the fact.

We sieved all subadult samples through stacked 1 mm (first) and 63 μm sieves (juveniles were not sieved due to their size) and then vacuum filtered the remaining liquid by pouring the solution onto a 5 μm Whatman polycarbonate filter using a Büchner funnel. A separatory funnel was used for most samples to ensure liquid was being poured in the center of the filter. Any large item on the sieves were looked at under a scope, and everything else was vacuum filtered. For the subadult fish, we filtered 100–300 mL per filter. Alcojet was used as needed to break apart lipid heavy digestate (Crichton et al., 2017). As juvenile fish samples were stored in glass jars, the jars and caps were rinsed and the resulting RO water poured onto the vacuum filter as well, resulting in 5 to 25 mL of filtered volume per sample per filter. We imaged all suspected plastics using a Leica EZ4W (camera included) or Leica EZ4 microscope equipped with a Motic 3.0 MP camera based on guidelines from Rochman et al. (2019) and Lusher et al. (2020). All but one of the microparticles that were run with FTIR were measured using the Leica EZ (mm), LevenhukLite software (mm), or “Motic Images Plus 3.0 ML” (µm), to ensure they were in the microparticle size range (<5 mm in one direction). The microparticles from juveniles were measured by one author, and the microparticles from the subadult fish were measured by another. All microparticles from subadults were measured in mm but converted to um. Despite the accidental exclusion of this single piece from being measured, it was retained in the results.

We followed protocols from Rochman et al. (2019) to determine morphology and color of the microparticles during this process as accurately as possible. Most of the filters were picked for microparticles for 1–3 h on average for the subadult fish at a later date. Once picked, suspected plastics (e.g., anthropogenic cellulosic, natural, synthetic, anthropogenic unknown) were placed into a verified polystyrene (marketed as acrylic) container or onto a double-sided piece of tape on a piece of projection article. We visually inspected these to determine if they could be plastic by probing them to see if they broke. If they broke, we did not count them as potential microparticles for future FTIR analysis (Lusher et al., 2020), unless they had already been verified via FTIR. There were some instances where we picked a subset of the microparticles found on filters due to the similarities of the pieces or picked pieces under different light. The term “microparticle” along with the classification categories described below are used to help differentiate microparticles from those that are analytically confirmed to be microplastics (Miller et al., 2021; Harris et al., 2021a).

Fourier transform infrared spectroscopy

We analytically verified a subset (~30% of all microparticles) of all suspected plastics from fish and blanks (~30% per Brander et al., 2020) on a Thermo Fisher Nicolet iN5, smart iTX, and Nicolet iS20 FTIR (Waltham, MA, USA). The particles were subsampled to represent the most common particle types collected. All microparticles were measured with Attenuated Total Reflectance or reflectance, and then followed by analysis using micro-Attenuated Total Reflectance (µATR). OMNIC software was used to generate spectra and Open Specy was used to provide final identification of material type. ATR with a Diamond or Germanium tip was used to run scans (64–256 in subadult fish, 128 in juveniles) of each microparticle to generate the resulting spectrum. We atmospheric corrected each spectrum no matter when the background was collected. This was due to the location of the FTIR relative to air sources. We matched spectra using several libraries including a library from Primpke et al. (2018), one created by the NOAA Marine Debris Research Program at UNCW, libraries created in-house, as well as standards that came with the machine. However, we did not match all pieces against all libraries as libraries were formed as this study progressed.

When a spectrum is matched to a library in OMNIC, a correlational match is produced. Due to the low qualifying spectra obtained for some spectra in OMNIC, the spectra generated were subsequently verified using Open Specy (Cowger et al., 2021). OMNIC provides the top 10 correlational matches for each spectrum, while Open Specy provides 100, and Open Specy uses multiple libraries specific to plastic polymers. To accept the generated spectrum for this study in Open Specy, the top five matches had to be >70. For consistency, all microparticles analyzed through Open Specy were smoothed at level 3, with baseline corrections at level 8. We did not account for the Ge tip minimum at 675 cm⁻¹. We categorized the microparticles into four categories based on Open Specy results: anthropogenic unknown, anthropogenic cellulosic, natural, and synthetic based on characterizations from Miller et al. (2021) similar to: Harris et al. (2021a), Caldwell et al. (2022). However, the microparticle holding jar and net that were verified with FTIR were not checked with Open Specy.

Quality assurance

We used several control procedures to account for background contamination. We cleaned all glassware using soap and water, rinsed with DI and RO water, then rinsed with 70% ethanol, and wrapped in foil prior to being baked at 400 °C for 4 h. We stamped a Whatmann filter with a 12 × 12 box grid and placed it into a glass petri dish each day a fish sample was open under the hood. This allowed us to account for plastic pieces that potentially deposited from the air into our samples within the hood. Our laminar flow hood was set up to allow computer access outside of the sample workspace, to reduce the potential for contamination.

We employed several blanks throughout the different microparticle processes. We used filtered and unfiltered DI, Milli-Q, and RO water (water source was switched mid-project due to lab move) to make solutions and to wet filters. To account for any contamination from water, we collected and vacuumed samples following the above procedures. We also used air blanks and KOH procedural blanks to track contamination throughout the process; however, we only included a subset in the final analysis. Finally, we followed recommendations to ensure minimal background contamination (Brander et al., 2020; Cowger et al., 2020). This included all project personnel wearing 100% cotton lab coats, mostly 100% cotton face masks (due to the COVID-19 pandemic), clothing as close to 100% cotton as possible, and nitrile gloves. Additionally, the container used to hold some microparticles and the mesh nets used for catching the juveniles were verified via µFTIR using 128 scans and a Germanium crystal (Fig. S2).

To determine the average number of microparticles attributed to background contamination (air, collection gear, personal accessories), the number of microparticles in the gastrointestinal tract was divided by the number of microparticles found in controls or blanks (air, KOH, water) and then multiplied by the number of controls divided by the number of fish (see Supplemental data), resulting in the number of microparticles per fish.

Study limitations

Differences in microparticle counts between fish size classes could be due to the fish size difference itself, but it is important to note that ingestion studies typically provide a snapshot view of microparticle consumption. Differences in the timing of last feeding, relative abundance of microparticles when feeding, clearance rates of microplastics from the gut, and even relative fish abundance (i.e., potential competition for prey items) could potentially affect instantaneous views of microparticles in the digestive tracts of fish. Athey et al. (2020b) found that polyethylene microspheres were retained within larval digestive tracts of silversides (Menidia beryllinia) up to 72 h after feeding, but we do not know the timing of passage for sub-adult and juvenile Black Rockfish. Given that the digestive period is relatively brief in most fish species, we have likely underestimated microparticles including microplastic consumption in Black Rockfish.

Our study, like many others (i.e., Jamieson et al., 2019; Ibrahim et al., 2020; Li et al., 2020a; Ragusa et al., 2021; Caldwell et al., 2022), did not include a recovery experiment. Recovery studies should be completed for experiments in the future to understand the success of the methods used. Additionally, this test could have provided beneficial information for understanding whether the concentration of KOH and temperature used in the oven degraded particles. Finally, our study is highly dependent on microscopy, as it is one of the first steps in particle recovery. Our FTIR is a manual machine, and therefore we needed to pick each individual particle that was a microparticle. Kotar et al. (2022) shows that microscopy is not the most effective method for microplastic analysis. With this in mind, we most likely underestimated the number of particles, especially as they decreased in size.

Data analysis

We conducted all statistical analyses in the R software environment using several libraries (car, carData, CARS, ggplot, multcomp, cowplot, mapdata, mapplots, maptools) and Excel. To determine if there was a statistically significant difference between the sampling sites and to account for differences in sample sizes, we conducted a generalized linear model (GLM) with Poisson error and log link using a GLM with Poisson error and a Tukey post-hoc test to identify any differences in the total number of pieces per fish across sites, at an alpha of 0.05. Wilson’s Score Intervals were used to find intervals for fish containing microparticles by site (Table S1). Additionally, confidence intervals can be found in Tables S4 and S5. Raw data and code can be found at: https://github.com/Brander-Harper/Black-Rockfish-Manuscript.

Impacts of anthropogenic pieces

Microparticles across all types were consumed by Black Rockfish, at an average of 7.31 (average background = 5) microparticles per subadult fish and an average of 4.21 (average background = 1.8) microparticles per fish for the juveniles (5.56 for all fish), with the majority (86.4%) of the verified polymer types being of unknown origin or cellulose based. A potential source of some of these anthropogenic or natural fibers could be from rope in the Dungeness fishery in Oregon. The material used for the twine on the pots for the largest commercial fishery in the state is cotton (Oregon Deparment of Fish & Wildlife, 2021). Our results are similar to Rochman et al. (2015), in which anthropogenic (albeit unverified) pieces were more prevalent than plastic in fish from the USA. Our results align with their finding that 12 study species were exposed to more textile fragments than synthetics, including Pacific Oysters (Crassostrea gigas), Yellowtail Rockfish (Sebastes flavidus), Chinook Salmon (Oncorhynchus tshawytscha), Vermilion Rockfish (Sebastes miniatus), Blue Rockfish (Sebastes mystinus), Pacific Mackerel (Scomber japonicus), Striped Bass (Morone saxatilis), Copper Rockfish (Sebastes caurinus), Lingcod (Ophiodon elongatus), Jacksmelt (Atherinopsis californiensis), Albacore Tuna (Thunnus alalunga), Pacific Anchovy (Engraulis mordax), and Pacific sanddab (Citharichthys sordidus). Due to the recurring presence of anthropogenic pieces outside of plastic polymers, a better understanding of their impacts on the health of organisms and humans is needed to assess their potential risk. Twenty-one Black Rockfish contained verified synthetics in our study, which is 17%; however, in Amorim, Ramos & Nogueira Júnior (2020) 12.4% of the fish contained plastics, but these pieces were not verified. Ingestion in our study is much higher than the 0.03% of fish that contained plastics examined by Steer et al. (2017) and similar to results found by Arias et al. (2019). Arias et al. (2019), showed that there were 241 non-verified microplastics found in 20 (12.05%) fish. Standard guidelines and methods are necessary to improve reproducibility and the ability to trace microparticles back to their source (Cowger et al., 2020).

Due to the complexities involved in verifying microparticle composition, some studies do not verify the material type using analytical verification (Rochman et al., 2015; Arias et al., 2019; Amorim, Ramos & Nogueira Júnior, 2020), but many studies such as ours now do (i.e., µFTIR). However, it is important to that note microparticles that are semi-synthetic or natural still need to be examined, as they are frequently detected (Athey & Erdle, 2021) and analytical identification is needed (Rochman et al., 2015; Steer et al., 2017; Alomar et al., 2017; Rochman et al., 2019; Allen et al., 2019; Brander et al., 2020; Cowger et al., 2020) to differentiate natural and cellulosic materials from synthetic or anthropogenically modified fibers, as they can also be quite similar visually (Athey & Erdle, 2021). There are also several different definitions that are used to classify microparticles that are neither plastic nor natural. Some of these include anthropogenic cellulosic, unknown potentially rubber, anthropogenic unknown, and unknown (Miller et al., 2021). Like the current lack of standardized analytical methodology, there is currently no standard terminology in use for microparticles that are not either clearly plastic or natural. Due to these complexities, we used both OMNIC (Thermo Fisher) and Open Specy (Cowger et al., 2021) to verify our spectral matches. Future analysis using FTIR could be improved upon by splitting the microparticles when possible before analysis to expose un-weathered or rugose surfaces, as this can produce increased FTIR matches (Jung et al., 2018). While the microplastics field has matured, the continued lack of standardization suggests the need for better harmonized methodologies for easier identification of microplastics and improved consistency (Brander et al., 2020; Cowger et al., 2020).

We sought to standardize methods across all of our analyses, and there were differences in the number of background microparticles detected for the different fish size classes; subadult fish had an average of five while juveniles had an average of 1.82. This could be due to the timing of when the samples were processed, how samples were stored and collected, or that the samples were processed in two different hood configurations. Regardless, this emphasizes the importance of these controls in the interpretation of microparticle data.

Toxicology

We did not measure the toxicity of the plastics in our fish. However, a number of studies have shown that reproduction is impacted when fish are exposed to different types of plastics at different sizes and concentrations (e.g., Chisada, Yoshida & Karita, 2019; Zhu et al., 2020), though some have not (Jacob et al., 2020). Additionally, plastics have been shown to cause cellular changes in fishes (e.g., (Karami et al., 2016; Jabeen et al., 2018; Espinosa, Esteban & Cuesta, 2019; Ahrendt et al., 2020; Jacob et al., 2020)) Fibers, specifically, have been used in toxicology studies to determine the impact of internalization and are known to impact body condition and to cause cellular changes in goldfish (Carassius auratus) (Jabeen et al., 2018) and respiration impacts in Black Sea Bass (Centropristis striata) (Stienbarger et al., 2021). Additionally, Amorim, Ramos & Nogueira Júnior (2020), saw that fibers, along with other morphologies, decreased Fulton’s K in subadults of Stellifer brasiliensis and hepatosomatic index in Micropogonias furnieri (Arias et al., 2019). Microplastics have been shown to affect a variety of endpoints in marine organisms, therefore, it is suspected that they can affect population stability. However, additional studies are needed to better understand the impact of microplastics on both population stability and ecosystem stability (Baechler et al., 2020b).

Fibers or filaments have been detected in several human samples including, but not limited to, colectomy (Ibrahim et al., 2020) and lung tissue (Jenner et al., 2022), however in a study on stool samples, fibers were not commonly found (Schwabl et al., 2019). Additional studies have looked at plastics in humans but did not look at morphologies and/or find fibers (e.g., (Ragusa et al., 2021; Leslie et al., 2022)).

Like many labs, we could not detect particles <20 um. Most labs have difficulty with this as seen in a multi-laboratory comparison study led by the Southern California Coastal Water Research Project (De Frond et al., 2022). In a article written by De Frond et al. (2022), the authors explain that the participants in the study (including our lab) had a recovery for particles >20 um at 92%. However, when considering particles <20 um, it is 32%. Since we do not have the ability to collect and then verify pieces that are this small, we risk missing many micro- and nanoparticles that can be consumed by marine life and potentially cause toxic effects (e.g., Cunningham et al., 2022; Siddiqui et al., 2022).

Spatial distribution

There are significant differences by location for microplastic ingestion by subadult Black Rockfish. ODFW samples from the two northern-most areas, near Cape Falcon and Cascade Head, had significantly higher levels of microparticle ingestion in comparison to those caught near Newport, Cape Perpetua and Redfish Rocks (Table S2). The lack of microparticle ingestion difference by juvenile Black Rockfish between Otter Rock marine reserve and Cape Foulweather is not surprising, as these two locations are physically close to one another (~3 nautical miles) (Oregon Department of Fish & Wildlife, Marine Resources Program, 2014). Although more data are needed to determine why there was a statistical difference in microparticle ingestion among sites, one explanation could be the northernmost sampling sites are in closer proximity to the Columbia River. Rivers can concentrate plastics across a broad terrestrial region that include dense human population centers away from the coast (Harris et al., 2021b), and the Columbia River, which flows though many municipalities including the Portland Metropolitan area, is, known to be contaminated with microplastics and other particle types but predominantly fibers (Talbot et al., 2022).

Our work suggests that fish located inside the reserves contain a similar amount of microparticles in both juvenile and subadult Black Rockfish, although we only compared juvenile Black Rockfish inside a comparison area to one marine reserve (Otter Rock, Oregon Department of Fish & Wildlife, Marine Resources Program (2021)). While ODFW subadult fish samples were not caught inside the reserves, the oceanographic processes known to influence comparison areas where samples were taken and reserves are similar (ODFW, 2022), suggesting similar microparticle exposure levels should be expected. This assumes that exposure to microparticles occurs mainly due to oceanographic processes and not from local land-based, inputs. Finally, it is of particular interest that there are several reasons to create and maintain marine reserves (White et al., 2010; Abessa et al., 2018) and many do not include protection from pollution, as was the case in Oregon. In many cases marine reserves have been placed in areas that are pre-contaminated with chemicals (Abessa et al., 2018) or are unknowingly in the pathway of newly emerging contaminants such as microplastics and other microparticles. Future marine reserve or protected area planning may benefit from considering locations of potential microplastic sources. This will enable careful consideration of reserve location, especially in coastal areas with both terrestrial and aquatic sources of plastics.