Microplastics and Anthropogenic Particles in Recreationally Caught Freshwater Fish from an Urbanized Region of the North American Great Lakes

Background: Microplastics are a pervasive contaminant cycling through food webs—leading to concerns regarding exposure and risk to humans. Objectives: We aimed to quantify and characterize anthropogenic particle contamination (including microplastics) in fish caught for human consumption from the Humber Bay region of Lake Ontario. We related quantities of anthropogenic particles to other factors (e.g., fish size) that may help in understanding accumulation of microplastics in fish. Methods: A total of 45 samples of six fish species collected from Humber Bay in Lake Ontario near Toronto, Ontario, Canada, were examined for anthropogenic particles in their gastrointestinal (GI) tracts and fillets. Using microscopy and spectroscopy, suspected anthropogenic particles were identified and characterized. Results: We observed anthropogenic particles in the GI tracts and fillets of all species. Individual fish had a mean±standard deviation of 138±231 anthropogenic particles, with a single fish containing up to 1,508 particles. GI tracts had 93±226 particles/fish (9.8±32.6 particles/gram), and fillets had 56±61 particles/fish (0.5±0.8 particles/gram). Based on a consumption rate of 2 servings/week, the average yearly human exposure through the consumption of these fish fillets would be 12,800±18,300 particles. Discussion: Our findings suggest that consumption of recreationally caught freshwater fish can be a pathway for human exposure to microplastics. The elevated number of particles observed in fish from Humber Bay highlights the need for large-scale geographic monitoring, especially near sources of microplastics. Currently, it is unclear what the effects of ingesting microplastics are for humans, but given that recreationally caught freshwater fish are one pathway for human exposure, these data can be incorporated into future human health risk assessment frameworks for microplastics. https://doi.org/10.1289/EHP13540


Table of Contents
Table S1.Inventory of collected fish samples from the Humber Bay region of Lake Ontario.Includes information on trophic level, sample size (n), sample mass (g), fish length (total length, cm).Trophic level was sourced from Fishbase, an online database which provides information on relative positions of fish within the food web based on the food items they consume.S3.Summary of microplastic particles observed in laboratory blank samples of gastrointestinal tracts by particles characteristics and the calculated mean, standard deviation (SD), and limit of detection (LOD).Note: B1, B2 etc., are sample IDs for each of the blanks.Table S4.Summary of microplastic particles observed in laboratory blank samples of fillets by particles characteristics and the calculated mean, standard deviation (SD), and limit of detection (LOD).Note: B1, B2 etc., are sample IDs for each of the blanks.Table S5.Percent recoveries from spike and recovery tests performed using both gastrointestinal (GI) tract and fillet sample procedures.Fillet procedure spike samples were spiked with 20 particles and then split into two equal subsamples.Thus, the number of particles recovered is sometimes greater than 10, the number expected if half of the particles were in each recovery sample.The percent recovery is based on recovering 10 particles, i.e., half of the particles spiked to test the recovery relevant to the subsampling and extraction and quantification procedure.Note: for the Sample IDs, SR stands for spike and recovery; PET, Polyethylene Terephthalate; PP, Polypropylene, PE, Polyethylene.Table S6.Number of suspected microplastic particles observed in each split sample, organized by particle morphology.Sample IDs are in the format of a 2-letter short form for species (SM = smallmouth bass, LM = largemouth bass, NP = northern pike, BB = brown bullhead, WS = white sucker) followed by a fish identifier ID number then 1/2 or 2/2 to differentiate the split samples created from the same fish.Note: g, grams; cm, centimeter.Table S7.Covariance results from tests of splitting methods using GI tract samples.Sample IDs are in the format of a 2-letter short form for species (SM = smallmouth bass, LM = largemouth bass, NP = northern pike, BB = brown bullhead, WS = white sucker).Note: g, grams; cm, centimeter.

Figure S1.
Flow chart of methods used to process GI tract and fillet samples from dissections to spectral analysis.

Figure S3
. Size histograms of all particle lengths (µm) after blank correction.(a) for particles 14 -5,000 µm in the gastrointestinal (GI) tract, (b) 5,000 -75,900 µm in the GI tract, and (c) all particles in the fillets.Corresponding data are presented in Excel Table S3.S3.

Figure S5.
Comparison of microplastic contamination between GI tract and fillet.Adjusted R 2 and p-value are indicated on the plot.Sample sizes for each species were n = 10 for brown bullhead (BB), n = 12 for largemouth bass (LB), n = 10 for northern pike (NP), n = 1 for rock bass (RB), n = 6 for smallmouth bass (SB), and n = 6 for white sucker (WS).Corresponding data are presented in Excel Table S1.

Figure S6
. Linear regression analysis between the log-transformed number of anthropogenic particles observed in the fillets (counts are standardized by grams of wet weight of the fillet tissue) and trophic level.Adjusted R 2 and p-value are indicated on the plot.Sample sizes for each species were n = 10 for brown bullhead, n = 12 for largemouth bass, n = 10 for northern pike, n = 1 for rock bass, n = 6 for smallmouth bass, and n = 6 for white sucker.Corresponding summary data are presented in Excel Table S1.

Figure S7
. Estimated intake of anthropogenic particles from the consumption of a single serving of fish (227 g) caught from the Humber Bay region of Lake Ontario, Canada.Sample sizes for each species were n = 10 for brown bullhead (BB), n = 12 for largemouth bass (LB), n = 10 for northern pike (NP), n = 1 for rock bass (RB), n = 6 for smallmouth bass (SB), and n = 6 for white sucker (WS).Note: g, grams.Corresponding summary data are presented in Excel Table S1.

Figure S8
. Estimated yearly intake of anthropogenic particles based on the consumption of two servings of fish weekly (227 g each serving, scaled up to a total of 23,608 g of fish for the year) caught from Lake Ontario, Canada.Sample sizes for each species are shown in brackets on the xaxis for northern pike (NP), brown bullhead (BB), largemouth bass (LB), rock bass (RB), smallmouth bass (SB), and white sucker (WS).Corresponding summary data are presented in Excel Table S1.Note: No., number; g, gram.Corresponding summary data are presented in Excel Table S1.

Literature Cited Additional File-Excel Document
Table S1.Inventory of collected fish samples from the Humber Bay region of Lake Ontario.Includes information on trophic level, sample size (n), sample mass (g), fish length (total length, cm).Trophic level was sourced from Fishbase, an online database which provides information on relative positions of fish within the food web based on the food items they consume.S3.S1.S1.

Figure S7
. Estimated intake of anthropogenic particles from the consumption of a single serving of fish (227 g) caught from the Humber Bay region of Lake Ontario, Canada.Sample sizes for each species were n = 10 for brown bullhead (BB), n = 12 for largemouth bass (LB), n = 10 for northern pike (NP), n = 1 for rock bass (RB), n = 6 for smallmouth bass (SB), and n = 6 for white sucker (WS).Note: g, grams.Corresponding summary data are presented in Excel Table S1.

Figure S2 .
Figure S2.Material type and chemical identification of suspected anthropogenic sample particles, blank particles, and suspected natural particles.a) relative abundances of material types of suspected anthropogenic particles based on Raman spectroscopy and µ-Fourier Transform Infrared (µFTIR) in samples and laboratory blanks; b) relative abundances of material types of suspected natural particles based on Raman spectroscopy and µFTIR; c) relative abundances of plastic polymers based on Raman spectroscopy and µFTIR in samples and laboratory blanks.Material type categories are based on Munno et al. 2021.Note: PE, Polyethylene; PEI, Polyetherimide; PP, Polypropylene; PS, Polystyrene; PET, Polyethylene Terephthalate; PVC, Polyvinyl Chloride; PA, Polyamide; PU, Polyurethane; ABS, Acrylonitrile Butadiene Styrene; PMMA, Poly(methyl methacrylate).Corresponding data are presented in Excel Table S2 and S3.

Figure S4 .
Figure S4.Non-metric multidimensional scaling (nMDS) showing arrangements of particles.(a) for size fractions between tissues; (b) for morphologies between tissues.P-values from the PERMANOVA and stress values for the nMDS are shown on the plot.Blue, filled circles represent the fillet samples, red, open squares represent the GI tract samples.Corresponding data are presented in Excel TableS3.

Figure S1 .
Figure S1.Flow chart of methods used to process GI tract and fillet samples from dissections to spectral analysis.

Figure S2 .
Figure S2.Material type and chemical identification of suspected anthropogenic sample particles, blank particles, and suspected natural particles.a) relative abundances of material types of suspected anthropogenic particles based on Raman spectroscopy and µ-Fourier Transform Infrared (µFTIR) in samples and laboratory blanks; b) relative abundances of material types of suspected natural particles based on Raman spectroscopy and µFTIR; c) relative abundances of plastic polymers based on Raman spectroscopy and µFTIR in samples and laboratory blanks.Material type categories are based on Munno et al. 2021. 2 Note: PE, Polyethylene; PEI, Polyetherimide; PP, Polypropylene; PS, Polystyrene; PET, Polyethylene Terephthalate; PVC, Polyvinyl Chloride; PA, Polyamide; PU, Polyurethane; ABS, Acrylonitrile Butadiene Styrene; PMMA, Poly(methyl methacrylate).Corresponding data are presented in Excel Table S2 and S3.

Figure S3 .
Figure S3.Size histograms of all particle lengths (µm) after blank correction.(a) for particles 14 -5,000 µm in the gastrointestinal (GI) tract, (b) 5,000 -75,900 µm in the GI tract, and (c) all particles in the fillets.Corresponding data are presented in Excel TableS3

Figure S4 .
Figure S4.Non-metric multidimensional scaling (nMDS) showing arrangements of particles.(a) for size fractions between tissues; (b) for morphologies between tissues.P-values from the PERMANOVA and stress values for the nMDS are shown on the plot.Blue, filled circles represent the fillet samples, red, open squares represent the GI tract samples.Corresponding data are presented in Excel TableS3.

Figure S5 .
Figure S5.Comparison of microplastic contamination between GI tract and fillet.Adjusted R 2 and pvalue are indicated on the plot.Sample sizes for each species were n = 10 for brown bullhead (BB), n = 12 for largemouth bass (LB), n = 10 for northern pike (NP), n = 1 for rock bass (RB), n = 6 for smallmouth bass (SB), and n = 6 for white sucker (WS).Corresponding data are presented in Excel TableS1.

Figure S6 .
Figure S6.Linear regression analysis between the log-transformed number of anthropogenic particles observed in the fillets (counts are standardized by grams of wet weight of the fillet tissue) and trophic level. 1 Adjusted R 2 and p-value are indicated on the plot.Sample sizes for each species were n = 10 for brown bullhead, n = 12 for largemouth bass, n = 10 for northern pike, n = 1 for rock bass, n = 6 for smallmouth bass, and n = 6 for white sucker.Corresponding summary data are presented in Excel TableS1.

Figure S8 .
Figure S8.Estimated yearly intake of anthropogenic particles based on the consumption of two servings of fish weekly (227 g each serving, scaled up to a total of 23,608 g of fish for the year) caught from Lake Ontario, Canada.Sample sizes for each species are shown in brackets on the x-axis for northern pike (NP), brown bullhead (BB), largemouth bass (LB), rock bass (RB), smallmouth bass (SB), and white sucker (WS).Corresponding summary data are presented in Excel TableS1.Note: No., number; g, gram.Corresponding summary data are presented in Excel TableS1.

Table S2 .
Particle morphology descriptions adapted from Rochman et al. 2019 and Werbowski et al. 2021 used to visually identify suspected microplastic particles under a microscope.

Table S2 .
Particle morphology descriptions adapted from Rochman et al. 2019 and Werbowski et al. 2021 used to visually identify suspected microplastic particles under a microscope.

Table S3 .
Summary of microplastic particles observed in laboratory blank samples of gastrointestinal tracts by particles characteristics and the calculated mean, standard deviation (SD), and limit of detection (LOD).Note: B1, B2 etc., are sample IDs for each of the blanks.

Table S4 .
Summary of microplastic particles observed in laboratory blank samples of fillets by particles characteristics and the calculated mean, standard deviation (SD), and limit of detection (LOD).Note: B1, B2 etc., are sample IDs for each of the blanks.

Table S5 .
Percent recoveries from spike and recovery tests performed using both gastrointestinal (GI) tract and fillet sample procedures.Fillet procedure spike samples were spiked with 20 particles and then split into two equal subsamples.Thus, the number of particles recovered is sometimes greater than 10, the number expected if half of the particles were in each recovery sample.The percent recovery is based on recovering 10 particles, i.e., half of the particles spiked to test the recovery relevant to the subsampling and extraction and quantification procedure.Note: for the Sample IDs, SR stands for spike and recovery; PET, Polyethylene Terephthalate; PP, Polypropylene, PE, Polyethylene.

Table S6 .
Number of suspected microplastic particles observed in each split sample, organized by particle morphology.Sample IDs are in the format of a 2-letter short form for species (SM = smallmouth bass, LM = largemouth bass, NP = northern pike, BB = brown bullhead, WS = white sucker) followed by a fish identifier ID number then 1/2 or 2/2 to differentiate the split samples created from the same fish.Note: g, grams; cm, centimeter.

Table S7 .
Covariance results from tests of splitting methods using GI tract samples.Sample IDs are in the format of a 2-letter short form for species (SM = smallmouth bass, LM = largemouth bass, NP = northern pike, BB = brown bullhead, WS = white sucker).Note: g, grams; cm, centimeter.

Table S1
. Note: No., number; g, gram.Corresponding summary data are presented in Excel TableS1.