Polyurethane microplastics and associated tris(chloropropyl)phosphate additives both affect development in larval fathead minnow Pimephales promelas

Abstract

Microplastics (< 5 mm) are a diverse class of contaminants ranging in morphology, polymer type, and chemical cocktail. Microplastic toxicity can be driven by one or a combination of these characteristics. Most studies, however, evaluate the physical effect of the most commercially available polymers. By disregarding other polymers with high consumption and/or production rates, and the chemical constituents of plastics, we fail to have a holistic understanding of the mechanisms of toxicity. Polyurethane is understudied in terms of effects testing yet considered one of the most hazardous polymers due to its chemical composition. Polyurethane is a high production polymer and is found in common consumer products ranging from packaging to spray foam insulation. To better understand the physico-chemical effects of polyurethane and a common additive in polyurethane products, we exposed larval fathead minnows for 28 days to polyurethane without chemical additives (i.e., plastic treatment), chemical leachate from polyurethane containing chemical additives (i.e., tris(chloropropyl)phosphate [TCPP]; i.e., chemical treatment) and polyurethane with chemical additives (i.e., plastic with chemical treatment) in a fully factorial experiment. We observed significant decreases in growth at 12 days posthatch (dph) in the plastic, chemical, and plastic with chemical treatments, suggesting a physical and chemical driver of toxicity. At 28 dph, we did not observe significant differences in growth, suggesting individuals can recover. We also observed concentrations of ΣTCPPs in fathead minnow exposed to the plastic with chemical treatment and the chemical only treatment, demonstrating TCPP uptake in exposed individuals. Combined, our data suggests the importance of both the physical and chemical components of microplastics when assessing effects, and thus emphasizing the need to evaluate the effects of microplastics in a multidimensional way.

Introduction

Freshwater ecosystems are under pressure from several anthropogenically induced stressors (Ormerod et al., 2010; Perkins et al., 2010) including climate change, fishery exploitation, invasive species, and pollutant loading (Schulte, 2014). Despite occupying less than 1% of the earth’s surface, freshwater ecosystems are biodiversity hotspots, containing 6% of all known species, and one-third of known vertebrate species (Hawksworth & Kalin-Arroyo, 1995). These biologically rich ecosystems are also considered sentinels (e.g., the canary in the coal mine) for anthropogenic perturbations (Perkins et al., 2010; Woodward et al., 2010) because they often experience higher loads of anthropogenic pollutants compared with marine systems.

Given its environmental ubiquity and evidence demonstrating adverse effects, plastic pollution has been identified as a contaminant of emerging concern (Lambert et al., 2014; Lambert & Wagner, 2018; Sutherland et al., 2010). Plastic emissions are expected to increase, even under the most conservative scenarios, due to high consumer demand and the mismanagement of waste (Borrelle et al., 2020). Microplastic research began by assessing the abundance in marine ecosystems with the earliest studies published in the 1970s (Carpenter et al., 1972; Colton et al., 1974). More recently, studies have begun to focus on quantifying microplastics in freshwater systems (González-Pleiter et al., 2020; Grbić et al., 2020; Rochman et al., 2022; Shruti et al., 2019; Talbot & Chang, 2022) and thereby highlighting their pervasiveness in aquatic ecosystems.

Experimental work suggests that microplastics (particles < 5 mm in size; [Arthur et al., 2009]) can affect organisms at multiple levels of biological organization (Bucci et al., 2020), including at the community (e.g., biodiversity, species composition; Green, 2016; Redondo-Hasselerharm et al., 2020), population (e.g., abundance; Bosker et al., 2019), individual (e.g., survival, growth; Bucci, 2022; Silva et al., 2019), and suborganismal (e.g., inflammation, oxidative stress; Qiao et al., 2019) levels. Still, there is much we do not know when it comes to the effect mechanisms of microplastics (Thornton Hampton et al., 2022). Microplastics are a diverse contaminant (Rochman et al., 2019) composed of various sizes, shapes, polymer types, and chemical additives. Most microplastic toxicity tests within the literature have focused on single stressor exposures by dosing one species to a uniform shape, polymer, and size (Paul-Pont et al., 2018). Studies often focus on polymer types that are most easily accessible for research, namely polyethylene (PE) and polystyrene (PS) and omit other polymer types that are commonly found in the environment, such as polyvinyl chloride (PVC) and polyurethane (PlasticsEurope, 2022; Zimmermann et al., 2020).

It is hypothesized that some microplastics may be more harmful than others (Lithner et al., 2011; Thornton Hampton et al., 2022). The physico-chemical properties of microplastics, such as size, shape, polymer type, and additive compounds, can influence their bioavailability and toxicity (Campanale et al., 2020; Verla et al., 2019). Plastic additives that provide specific properties, such as flexibility, color, UV protection, and durability, (Hahladakis et al., 2018; Hermabessiere et al., 2017), can account for nearly 60% of the total mass of the material (Net et al., 2015) and often are not chemically bound to the polymeric matrix. Thus, additives can easily leach into the environment (Hermabessiere et al., 2017). Studies have suggested that there may be different biological effects from different polymer types due to their chemical constituents(Groh et al., 2023; Hahladakis et al., 2018; Zimmermann et al., 2019). For example, styrenic compounds that are often found in PS are known endocrine disruptors (Lithner et al., 2011), whereas chemical additives such as flame retardants or UV filters have been shown to have a range of toxicological effects, such as oxidative stress and cytotoxicity (Zimmermann et al., 2019). There are few studies that demonstrate the effects of plastic leachates (Delaeter et al., 2022); Hermabessiere et al., 2017), and even fewer assessing the combined toxicity of the physical and chemical components of microplastics ((Bucci et al., 2022; Rochman, 2013; Schrank et al., 2019; Zimmermann et al., 2019). There is a need to fill this knowledge gap by testing effects of different polymer types with different additive chemicals to better understand the biological impacts of microplastics (Barrick et al., 2021; Thornton Hampton et al., 2022).

Here, we measured the effects of a commercial polyurethane with chemical additives. Polyurethane is one of the most commonly produced plastics (PlasticsEurope, 2022) with a wide range of applications (e.g., paints, adhesives, fibers/textiles; [Shah et al., 2008]) and commonly found in the environment. Polyurethane has high chemical additive loads and has been suggested as one of the most hazardous plastic types due to its chemical composition (Lithner et al., 2011; Zimmermann et al., 2019). To tease apart the role of chemical additives on microplastic toxicity, we aimed to measure effects of polyurethane with and without chemical additives in Pimephales promelas (fathead minnow), as well as the chemical additive alone. Over a 28-day exposure period, we measured growth endpoints to determine single and combined effects of polyurethane and plastic additives. We also measured the concentrations of chemical additives in the fish after exposure to polyurethane with a chemical additive (i.e., tris(chloropropyl)phosphate [TCPP]) and a chemical additive alone. As seen with other polymers, we predicted we would observe negative effects to development among plastic treatments. We also predicted we would observe a difference in effects between the physical and chemical treatments due to different physical and chemical drivers of toxicity. Investigating the potential toxicity of the physio-chemical interface of different polymers and additives is vital in understanding the ecotoxicological effects; thus, helping inform robust, multidimensional risk assessments.

Methods

Preparation of microplastics

Our experimental design included polyurethane without chemical additives (i.e., referred to as the plastic treatment; Goodfellow Polyurethane Rod: model UR30-RD-000125 [additive free]) and polyurethane with chemical additives (i.e., with additives, referred to as the plastic with chemical treatment; Touch N Foam No Warp Window & Door Foam Sealant). The specific plastic was chosen because of its percent weight (10%–30%) of tris(2-chloro-1-methylethyl) phosphate (technical TCPP, three isomers; Chemical Abstracts Service no.: 26248-87-3), used as a flame-retardant, common in insulating spray foams (Babrauskas et al., 2012; van der Veen & de Boer, 2012). Tris(2-chloro-1-methylethyl) phosphate belongs to a large class of synthetic compounds called organophosphate esters (OPEs) and has been used as a substitute for brominated and other chlorinated flame retardants after their usage was restricted and/or banned (Blum et al., 2019; National Toxicology Program, 2020). Organophosphate esters are often used as flame retardants (OPFRs) and plasticizers in a wide variety of industrial applications (van der Veen and de Boer, 2012). Generally, OPFRs, are thought to be less persistent in the environment than other classes of flame retardants (e.g., polybrominated diphenyl ethers [PBDEs]; Zhang et al., 2016). Yet, some OPFRs, like TCPP, are environmentally abundant with elevated concentrations across the world (Rauert et al., 2018), which has been noted as a cause for concern (Blum et al., 2019).

For the plastic treatment, the polyurethane (without chemicals) was ground using a stainless steel file (milled-tooth cabinet file precleaned, with 1:1 acetone: hexane). For the plastic with chemical treatment, the polyurethane was ground using a stainless-steel coffee grinder (solvent-cleaned, Hamilton Beach). Although we aimed to use the same methods for grinding each plastic type, we were unable to due to their properties. The plastic treatment (i.e., polyurethane rod) was extremely rigid, and the plastic with chemical additives was foamy, light, and produced static. Both methods created irregularly shaped fragments. To isolate a size range of 100 to 500 µm, each ground plastic was sieved with stainless steel sieves (Bucci et al., 2022). The mass concentrations for each type of polyurethane were obtained by determining the mass-count ratio. Briefly, for each plastic type, three different masses of polyurethane were weighed on individual petri dishes. The particles in each petri dish were photographed and counted three times (See online supplementary material Table S1 for the masses and counts). A random subsample of 100 particles per treatment was also measured and particle sizes were not significantly different between treatments (t-test, p > 0.05). Then, we used the known mass and count to determine the mass needed to reach 20,000 particles. The resulting fragments used in the experiment ranged in size from 100 to 500 µm, which is representative of the diversity of fragments organisms are exposed to in the environment. More detail can be found in the online supplementary material (Tables S1 and S3 and Figure S1).

Animal care

Pimephales promelas (hereafter referred to as fathead minnow) eggs were obtained from breeding pairs at AquaTox (Puslinch, Ontario, Canada). Fish maintenance and sampling followed protocols approved by University of Toronto Animal Care Committee (AUP- 01166388). Eggs (< 24 hr postfertilization) were rolled off clay spawning tiles into prebaked (400°C for 7 hr) petri dishes. Eggs (40 per beaker) were randomly mixed into twenty-four 500 ml beakers, with pre-aerated, dechlorinated water. Once the eggs hatched (∼3 days postfertilization), the fish were fed once per day with newly hatched brine shrimp (Parrott, 2005). The beakers were held in water baths maintained at 24 ± 1°C with a 16:8-hr light:dark photo period. Temperature was monitored daily. Water quality measurements (dissolved oxygen [DO], conductivity, and ammonia) were taken twice a week from one random beaker in each water bath. Water quality measurements indicated favorable conditions throughout the 28-day exposure (i.e., averages ± SD as follows: DO = 7.72 ± 0.71 mg/L; pH = 7.38 ± 0.28; ammonia = 0.2 ± 0.26; temperature = 23.35 ± 0.55°C; see online supplementary material Table S4). If a beaker failed to meet the water quality parameters, a full water change of all beakers was conducted. This happened once during the experiment in a control beaker. Daily, 30% water changes occurred and fecal matter was removed using a soft nylon net. Full (100%) water changes with renewed plastic treatments occurred once per week.

Experimental design

To assess the physical and chemical effects of polyurethane and relevant chemical additives on larval fathead minnows, a 28-day posthatch (dph) exposure (adapted from Organisation for Econmic Co-operation and Development [OECD, 2013]) was conducted with four treatments (n = 6 vessels per treatment; Figure 1) starting at 0 dph. The experimental design was fully factorial with two factors: plastic and chemical additive. Each factor had two levels: zero contaminant and contaminant present. This resulted in organisms being exposed to either the physical particle only (plastic treatment), the chemical additive only (chemical treatment), the physical particle with chemical additive (plastic with chemical treatment), and no contaminants (control). The treatments were randomly arranged across two water baths with three replicates per treatment in each.

Figure 1.

Graphic outlining the fully factorial experimental design, including four treatments (i.e., control, polyurethane only, polyurethane with chemical, and chemical only) each with six replicates (n = 6).

The amount of microplastics used for each treatment consisted of a nominal dose of 20,000 particles/L. We chose a concentration that is higher than environmental concentrations (Koelmans et al., 2020), to help meet our objective of being able to tease apart whether effects were due to the physical particle and/or the chemical additive. The concentrations in mass per unit volume corresponded to 24.02 mg/L for polyurethane without chemical and 23.90 mg/L for polyurethane with chemical. Chemical treatments were prepared by adding the polyurethane with chemical additive (i.e., 23.90 g/L) to precleaned 60 µm Nytex and housed in a precleaned stainless steel tea-steeper. The tea-steeper and the Nytex were washed, and triple rinsed with dechlorinated water. The tea-steeper was suspended in the water column of the 1 L beaker, which allowed for active leaching throughout the exposure period (See online supplementary material Figure S2).

All treatments were renewed weekly with complete (100%) water changes to maintain the nominal dose. During the experiment, survival was monitored daily, and any morphological deformities present on deceased individuals were recorded. On Day 12 and Day 28 of exposure, individuals were culled with 250 mg/L of tricaine methansulfonate (MS-222) buffered with 250 mg/L of sodium bicarbonate. On 12 dph, 15 random individuals from each replicate were culled, photographed, and weighed as a pooled group. Individuals were stored in precleaned glass vials and due to their size at 12 dph, individuals were pooled by treatment to ensure enough mass for chemical analysis. On 28 dph, each remaining fish in each replicate vessel was culled, photographed, weighed, and stored in prebaked foil for chemical analysis. At 12 dph and 28 dph, the length of each fish was measured using ImageJ (Schneider et al., 2012). The quality of each larva was evaluated based on whether it was visibly deformed (Table 1). The deformities were classified as cranial (irregular shaped), focal (irregular shaped), scoliosis (spinal curvature), edema (fluid build-up around the eyes, heart, and/or yolk sac), tail truncation, or other (Bucci et al., 2022). Each deformity was counted individually such that fish with multiple deformities were counted multiple times. Because decisions about deformities can be subjective, this data is semiquantitative. See online supplementary material Figure S3 for examples of each type of deformity.

Table 1.

Number of observed deformities in fish exposed to plastic, chemical, and plastic with chemical treatment.

	Cranial	Focal	Edema	Scoliosis	Tail
12 dph
Control	0	0	0	0	0
VPUF	2	4	0	1	0
CMPU	0	1	1	0	0
LCHE	1	0	0	0	1
30 dph
Control	0	0	0	0	0
VPUF	0	0	0	0	0
CMPU	0	1	0	0	0
LCHE	0	2	0	0	0

Chemical analysis

At 12 dph, whole fathead minnows were pooled by treatment. At 28 dph, whole fathead minnows were pooled by beaker (i.e., replicate). Note that fish were not dissected prior to chemical analysis due to the small size of the fish. Thus, concentrations of additives include the digestive tract. Thus, chemicals in fish from treatments with physical particles may be biased higher due to microplastics with chemicals in the stomach. Whole fathead minnows were homogenized using a triple-rinsed solvent-cleaned mortar and pestle. Each sample was then sonicated in dichloromethane (DCM; Millipore). Following sonication, DCM was removed by pouring into a round bottom flask. Iso-octane (Millipore) was added as a keeper solvent and the sample was concentrated to 5 ml via Rotovap (Buchi). The extract was then dried over a column of sodium sulfate (granular, JT Baker), eluted into a test tube, and blown down to 1 ml with a gentle stream of nitrogen (Organomation N-EVAP). Lipids were removed using florisil (Supelco) eluted with ethyl acetate (EMD Science) and solvent exchanged into iso-octane. Mirex (100 ng, Accustandard) was added to each sample as the internal standard for time reference and volume correction factor and relative response calibration curves were created (See online supplementary material Figure S4). Extraction blanks (n = 3) followed the same method and were processed alongside the samples. Solvent blanks (n = 2) were also included where pure solvent was started at the florisil step. Concentrations in the solvent and extraction blanks were both very low so they averaged and subtracted from each sample. Analyses were conducted using an Agilent 8890 gas chromatograph coupled with an Agilent 7010B GC-triple quad mass spectrometry; analysis was conducted in single-quad mode. Commercially, TCPP is produced as a technical mixture (Truong et al., 2017); therefore, concentrations for total TCPP as well as the individual isomers found in the technical mixture are reported (Truong et al., 2017).

Statistical analysis

All statistical analyses were done using R, version 4.2.1 (RDevelopment Core Team 2021). To test for differences between treatments, we ran individual nested three-factor analysis of variance (ANOVA) tests with vessels nested in treatment for each individual endpoint where there were measurements for each fish, that is, length and weight at 28 dph (See online supplementary material Table S2). Where measurements were pooled across individuals, we used 2-factor ANOVAs (Factor 1: plastic; Factor 2: chemical), that is, survival, mass at 12 dph, and concentrations of chemicals at 28 dph. Where p < 0.05 was deemed significant. To ensure that the data fit the assumptions of an ANOVA, all data were tested for normality (Shapiro‐Wilk test) and homogeneity of variance (Bartlett’s test). The mass data at 28 dph failed the homogeneity of variance test. To meet the ANOVA assumptions, the 28 dph mass data were log-transformed. Survival data is a proportion and was logit transformed. Tris(chloropropyl)phosphate concentration data failed the homogeneity of variance test, including after log-transformation. Because ANOVAs are robust to deviances from normality and homogeneity of variances when experimental designs have even sample sizes across treatments (Underwood, 1997), we still ran parametric tests. When a factor was significant, a Tukey’s post hoc test was performed to determine which treatments were statistically different (p < 0.05).

Results and discussion

Effects of polyurethane on fathead minnows

We exposed larval fathead minnow to plastic, chemical, and plastic with chemical treatments in a fully factorial experiment with a negative control. The survival of the controls met or exceeded the test acceptability criteria, that is, > 75%, (OECD, 2013) with an average (± SD) survival of 94 ± 0.05% in the control, 96 ± 0.06% in the plastic treatment, 93 ± 0.04% in the chemical treatment, and 95 ± 0.035% in the plastic with chemical treatment. There were no significant differences in survival across treatments during the 28-day exposure (p > 0.05; see online supplementary material Table S2). Our results differed from Dinani et al. (2021), who observed a time- and dose- dependent increase in zebrafish mortality exposed to three size classes (e.g., < 100 µm, 100–300 µm, 300–1,000 µm) and concentrations of polyurethane particles ranging from 0–1000 µg/L (Dinani et al., 2021). It is unclear whether Dinani et al. (2021) used virgin or commercial polyurethane microplastics in their exposures. In the case of the latter, such high mortality rates could be explained by additive chemical toxicity. Like our study, Zimmermann et al. (2019) investigated polyurethane, with and without a chemical component, in Daphnia magna and also did not observe any significant differences in survival across treatments suggesting exposure to polyurethane particles may not result in acute lethality. Dinani et al. (2021) reports dose-dependent mortality with an median lethal dose of 10 µg/L of polyurethane particles < 100 µm, which supports the hypothesis that smaller particles may be more toxic due to an ability to translocate (Browne et al., 2008) and an increased surface area to volume ratio (Goedeck et al., 2017). Because particle mass data were not reported by Dinani et al. (2021), it is challenging to compare their work to Zimmermann et al. (2019) and ours presented here. Without more information, we cannot conclude what caused the discrepancy among the results of each experiment.

We observed sublethal effects relevant to development. Overall, growth (i.e., length and mass) varied across treatments but only at the first sampling point. At 12 dph, the average (± SD) length across treatments was 9.9 ± 0.5 mm in the control, 9.4 ± 0.5 mm in the plastic treatment, 9.6 ± 0.3 mm in the chemical treatment, and 9.4 ± 0.4 mm in plastic with chemical treatment (Figure 1B). The ANOVA showed that the factor plastic was significant (p < 0.05), with a significant difference between fish exposed to plastic and fish not exposed to plastic, that is, fish from the plastic and plastic with chemical treatments were smaller in length than fish from the control and chemical treatments (p < 0.05; See online supplementary material Table S2). The average (± SD) mass across treatments at 12 dph was 0.04 ± 0.002 g in the control, 0.03 ± 0.005 g in the plastic treatment, 0.04 ± 0.005 g in the chemical treatment, and 0.03 ± 0.004 g in plastic with chemical treatment (Figure 2A). The ANOVA showed a significant interaction between plastic and chemical (p < 0.01) for 12 dph mass (See online supplementary material Table S2). A post hoc test showed a significant difference between the exposure treatments and the control, with no difference among exposure treatments. The exposed fish in the plastic, chemical, and plastic with chemical treatments were significantly lighter than the fish in the control treatment (Figure 2A). By 28 dph, ANOVAs showed no significant differences between treatments for mass or length. Still, the patterns were similar to the 12 dph. The average (± SD) length was 15.7 ± 0.9 mm in the control, 15.0 ± 0.5 mm in the plastic treatment, 15.2 ± 0.4 in the chemical treatment, and 15.1 ± 0.2 in the plastic with chemical treatment (Figure 2D). The average (± SD) mass was 0.05 ± 0.2 g in the control, 0.03 ± 0.004 g in the plastic treatment, 0.05 ± 0.03 g in the chemical treatment, and 0.04 ± 0.01 g in the plastic with chemical treatment (Figure 2C). The lack of a significant difference suggests that individuals caught up in terms of development.

Figure 2.

Boxplots illustrating the impact of two factors, plastic and chemical, on the measured endpoints: 12 dph mass (A), 12 dph length (B), 28 dph mass (C), 28 dph length (D). Significance letters reflect the results of a Tukey’s post hoc test to determine significance among treatments following analysis of variance. Significance was determined when p < 0.05.

A range of deformities were observed in fish exposed to polyurethane and associated chemical additives. There were no deformities observed in the control, and 14 deformities observed across exposure treatments. Of the observed deformities, 78% were observed at 12 dph. At 12 dph, we observed seven deformities in the plastic treatment and two in each of the chemical and plastic with chemical treatments. Most deformities observed were focal abnormalities, for example, pin eyes (Figure 3B). We also observed cranial (blunt head) and tail truncation in the chemical treatment and edemas in plastic with chemical treatments (Table 1). At 28 dph, we observed three deformities in total, and only in the chemical and plastic w/chemical treatments (i.e., only in treatments where additive chemicals were present). All deformities at 28 dph were focal, with two in the chemical treatment and one observed in the plastic with chemical treatment (Table 1).

Figure 3.

Examples of deformities observed during the exposure period. A: edema and cranial abnormality; B: focal abnormality; C: scoliosis. Observed deformities are shown with red arrows.

We observed a reduction in mass at 12 dph in both the plastic, chemical, and plastic with chemical treatments, which suggests that changes in development may be caused by physical and chemical mechanisms. However, for length, the reduction in size was only observed in the treatments with physical particles present (plastic and plastic with chemical treatments). This discrepancy in results may be by chance, or relevant to different effect mechanisms affecting mass and length. Similarly, we observed deformities in the plastic treatment as well as in both treatments where additive chemicals were present (i.e., chemical and plastic with chemical treatments). Combined, these results suggest that there are both physical and chemical effects of polyurethane.

Across the literature, effects have been demonstrated from physical particles and chemical additives. Zimmermann et al. (2019) observed significant declines in reproductive output of D. magna in treatments with polyurethane particles with and without chemicals. Such effects may be due to a reduction in energy due to food dilution (Mehinto et al., 2022). Significant growth effects in particle-only treatments have been reported for other polymer types, including PS (Ziajahromi et al., 2017) and polyester (Walkinshaw et al., 2023). Further, Antonopoulou et al. (2022) observed reduced growth in freshwater algae (Chlorococcum sp.) exposed to TCPP only, demonstrating an effect of the additive (Antonopoulou et al., 2022). Schrank et al., (2019) observed significant negative growth effects of D. magna exposed to PVC with additives compared to PVC without an additive, similar to Zimmermann et al. (2019). Here, we observed changes in development relevant to both particles and chemicals. These discrepancies across the literature suggest differences in polymer and chemical makeup could lead to different observed effects and effect mechanisms.

The focal deformities observed in the chemical treatments, although semiquantitative, are aligned with observations of OPE induced deformities in the literature. For example, Alzualde et al. (2018) observed a significant reduction, and sometimes an absence, in eye size of zebrafish embryos exposed to another OPE, tris(1,3-dichloro-2-propyl) phosphate (Alzualde et al., 2018). Studies evaluating the effects of other plastic leachates have also observed focal abnormalities, such as pin eyes and a loss pigmentation in larval fathead minnow exposed to tire leachates (Chibwe et al., 2021). Further, structural cranial deformities could impact foraging behavior. Wirt et al. (2018) observed a 10% decrease in foraging efficiency, likely due to cranial deformities (i.e., blunt heads) in zebrafish exposed to triclosan (Wirt et al., 2018), an antimicrobial compound often added to plastic products (National Industrial Chemical Notification and Assessment Scheme, 2009; Scientific Committee on Consumer Safety, 2010). Deformities that impact foraging behavior, such as blunt heads and pin eyes, could lead to negative growth impacts as individuals age and mature.

Today, there is a gap in our understanding about how different plastic products may drive both physical and chemical toxicity. At present, microplastic shape and size are thought to be important drivers of toxicity (Thornton Hampton et al., 2022). In some studies, researchers show that more complex shapes (e.g., microfibers, fragments) have greater toxicity (Au et al., 2015; Ogonowski et al., 2016) and that smaller particles are more likely to translocate into tissues leading to potential bioaccumulation (Mehinto et al., 2022). Smaller particles also have a greater surface area to volume ratio, which may enhance leaching and thus bioavailability of associated chemicals (e.g., Hahladakis et al., 2018; Shen et al., 2019). There is a need for future work to design experiments that account for the complexity of microplastics and evaluate microplastics as a multidimensional contaminant. This includes investigating not only the effects of the physical particle (size, shape) and polymer type but also the chemical mixture affiliated with different polymer types (e.g., monomers, additives).

Concentrations of TCPP in fish

We observed a measurable body burden of TCPP in all treatments (Figure 4). At 12 dph, we observed ΣTCPP concentrations of 3,442 ng/g in the control, 552 ng/g in the clean plastic treatment, 6,605 ng/g in the chemical treatment, and 28,654 ng/g in the plastic with chemical treatment. Here, we only had one sample per treatment (i.e., individuals were pooled to ensure enough mass was present) so there is no statistical test. At 28 dph, the pattern was similar with an average (± SD) ΣTCPP of 26,934 ± 48,964ng/g in the control, 4,777 ± 5,792ng/g in the clean plastic treatment, 85,331 ± 29,321ng/g in the chemical treatment, and 1,816,245 ± 862,462ng/g in the plastic with chemical treatment. The ANOVA showed a significant interaction between plastic and chemical (p < 0.001). A post hoc test showed no significant difference between the control and the clean plastic treatment. The chemical treatment was significantly higher than the control and clean plastic treatment, and the plastic with chemical treatment was significantly higher than all three other treatments. It is difficult to discern the results of the plastic with chemical treatment due to the potential for plastic in the gut content in samples analyzed chemically; however, the chemical treatment demonstrates that additives leached from polyurethane can be taken up and concentrate in fish.

Figure 4.

Boxplots illustrating the impact of two factors, plastic and chemical, on total body burden of tris(chloropropyl)phosphate (TCPP) in fathead minnow at 28 dph. Significance letters reflect the results of a Tukey’s post hoc test to determine significance among treatments following analysis of variance. Significance was determined when p < 0.05.

Given that OPEs are a common chemical additive to a variety of consumer products, including plastics, there is a need to better understand the role microplastic particles have in facilitating the uptake and overall exposure of additive chemicals. Some studies have shown that the uptake of additives via microplastic is not as relevant as aqueous pathways. For example, Jang et al. (2021) compared PS ingestion and leachate uptake of hexabromocyclododecane in Mediterranean mussels (Mytilus galloprovincialis) and reported uptake of hexabromocyclododecane through surrounding water was a more significant pathway for chemical uptake than PS particle ingestion (Jang et al., 2021). Although OPEs, like TCEP (i.e., tris(2-carboxyethyl)phosphine hydrochloride) and TClPP (the dominant component of technical TCPP), have been found to accumulate in fish (Guo et al., 2017; McGoldrick et al., 2014) and herring gull eggs (Greaves et al., 2016; Greaves & Letcher, 2014) from the Laurentian Great Lakes, it is unclear the role microplastics have in the transport and exposure of OPEs. However, Sühring et al. (2022) observed relatively high concentrations of nonchlorinated OPEs in various Arctic seabird tissues (e.g., muscle, liver, brain) compared with chlorinated OPEs (TCPP is a Cl-OPE) and PBDEs demonstrating bioaccumulation in nature (Sühring et al., 2022). Further, Sühring et al. (2022) observed a significant relationship (p < 0.05) between ingested microplastics and both non-Cl-OPEs and Cl-OPEs in northern fulmars from Prince Leopold Island, suggesting microplastics may have a role as an exposure pathway of OPEs in nature.

Generally, OPEs have been termed as an “environmentally friendly” alternative to PBDEs due to their hypothesized lower environmental persistence and bioaccumulation potential. These ideas have become heavily scrutinized as laboratory studies and field observations (e.g., Ma et al., 2017; Sühring et al., 2022; Wang et al., 2017) challenge these hypotheses. Our observations of relatively high chemical burdens of TCPP in the chemical treatment, as well as an increase over time, suggests that TCPP could bioaccumulate. Therefore, future assessments should consider these compounds as bioavailable and further assess the role of microplastics as a source.

Conclusion

Microplastics are a physical and chemical stressor, likely due to the fact that they are a multidimensional contaminant (e.g., size, shape, chemistry). By exposing fathead minnow larvae to virgin and commercial polyurethane and polyurethane additives (i.e., TCPP), we show that effects can be due to the physical presence of particles and/or the chemical additives. At 12 dph, we observed decreases in growth in both physical and chemical treatments. However, these effects were not observed at 28 dph, suggesting that exposed individuals could recover over time. We also observed deformities in treatments with and without additives and/or particles present. Finally, we saw uptake of TCPP in fish exposed to additives. To better understand the effects of microplastics, with respect to their physical and chemical components, future work should focus on evaluating effects of different polymer types and also include different relevant additive chemicals and combinations of each. By assessing the multidimensionality of microplastics, we can work toward a more holistic understanding of ecotoxicological effects and thereby, inform ecological risk assessments.

Supplementary material

Supplementary material is available online at Environmental Toxicology and Chemistry.

Data availability

Data is provided in the supplementary material of this article.

Author contributions

Bonnie Hamilton (Conceptualization, Data curation, Formal analysis, Methodology, Project administration, Visualization, Writing—original draft, Writing—review & editing), Liisa Jantunen (Formal analysis, Resources, Writing—review & editing), Chelsea Rochman (Conceptualization, Project administration, Resources, Supervision, Writing—original draft, Writing—review & editing)

Funding

Department of Ecology and Evolutionary Biology, University of Toronto.

Conflicts of interest

The authors declare no conflicts of interest.