Toxicity of tire particle leachates on early life stages of keystone sea urchin species

Particles from tires are a major fraction of microplastic pollution. They contain a wide range of chemical additives that can leach into the water and be harmful to aquatic organisms. In this study, we investigated the acute toxicity of tire particle leachates in early life stages of three keystone echinoderm species ( Paracentrotus lividus , Arbacia lixula , Diadema africanum ). Embryos were exposed for 72 h to a range of leachate dilutions, prepared using a concentration of 1 g L (cid:0) 1 . Larval growth, abnormal development, and mortality were the measured endpoints. Furthermore, we estimated the activity of glutathione S transferase (GST) and the electron transport system (ETS) in P. lividus . Strong concentration-dependent responses were observed in all species, though with differing sensitivity. The median effect concentrations for abnormal development in P. lividus and A. lixula were 0.16 and 0.35 g L (cid:0) 1 , respectively. In D. africanum , mortality overshadowed abnormal development and the median lethal concentration was 0.46 g L (cid:0) 1 . Larvae of P. lividus were significantly smaller than the control from 0.125 g L (cid:0) 1 , while the other two species were affected from 0.5 g L (cid:0) 1 . ETS activity did not change but there was a non-significant trend of increasing GST activity with leachate concentration in P. lividus . Seven organic chemicals and eight metals were detected at elevated concentrations in the leachates. While we regard zinc as a strong candidate to explain some of the observed toxicity, it can be expected that tire particle leachates exhibit a cocktail effect and other leached additives may also contribute to their toxicity. Our results emphasize the importance of multi-species studies as they differ in their susceptibility to tire particle pollution. We found negative effects at concentrations close to projections in the environment, which calls for more research and mitigation actions on these pollutants.


Introduction
Global microplastic (MP) pollution has been researched extensively in the past decade, with most studies focusing on MP ingestion. However, recent research suggests that leached additives are the primary cause of MP toxicity in aquatic organisms Manzo and Schiavo, 2022). In plastic production, thousands of chemicals are used as monomers, additives, and processing aids, with the number and amount of additives varying depending on the type and application of the plastics. Wiesinger et al. (2021) identified more than 10,000 relevant chemicals, including over 2,400 of potential concern and around 4,100 for which no hazard classification is available. These substances are weakly bound to plastic polymers and can leach into the environment, where organisms may be exposed to them (Wiesinger et al., 2021). However, the effects of plastic-leached additives on biota and ecosystems remain poorly understood. Therefore, more research on the ecological impact of additives from different plastic materials is needed.
Tire wear particles (TWPs) have recently been identified as a major source of MPs (Baensch-Baltruschat et al., 2020;Rødland, 2019) with a potentially high risk of environmental impact . TWPs ☆ This paper has been recommended for acceptance by Maria Cristina Fossi. are generated by abrasion of tires against the road surface (Kole et al., 2017). Despite being mainly composed of natural and synthetic rubber polymers, elastomers like rubber have been largely ignored in MP research due to unclear definitions (Hartmann et al., 2019) and challenging characterization (Baensch-Baltruschat et al., 2020). Estimates suggest that almost six million tons of TWPs were released globally in 2016, constituting 5-10% of the global mass of plastics entering the oceans (Kole et al., 2017). Other studies indicate that TWPs are the major or one of the biggest sources of MPs in the environment (Hann et al., 2018;Lassen et al., 2015;Rødland, 2019). Besides rubber, TWPs are typically composed of carbon black, oils, plasticizers, vulcanizing agents and other additives (Kreider et al., 2010;Rodgers and Waddell, 2013;Wagner et al., 2018). Müller et al. (2022), identified 214 organic compounds in tire wear of which 145 were classified as leachable. Also, several studies found high concentrations of metals in TWP leachates (Capolupo et al., 2020;Halsband et al., 2020). While acute toxicity of TWP leachates as well as sub-lethal effects on behavior, development, feeding, reproduction, and biomarkers have been studied in copepods, rotifers, and bivalves (Capolupo et al., 2020;Halsband et al., 2020;Lehtiniemi et al., 2021;Shin et al., 2022;Tallec et al., 2022bTallec et al., , 2022aYang et al., 2022), the number of studied invertebrate species is still limited. Adverse effects of TWP leachates or the tire-derived compound 6PPD-quinone have also been reported for several fish species, resulting in impaired eye development, changes in swimming behavior and mortality (Brinkmann et al., 2022;Chang et al., 2023;Kim et al., 2023;McIntyre et al., 2021;Tian et al., 2021). However, it has also been demonstrated that sensitivity to tire-derived additives can be highly species-specific, such as for 6PPD-quinone (Brinkmann et al., 2022;Foldvik et al., 2022;Tian et al., 2021).
Sea urchins are crucial to coastal benthic communities, contributing to the resilience of coral reefs and shaping benthic rocky communities in temperate habitats through their grazing activity (Agnetta et al., 2013;Dang et al., 2020;Pearse, 2006). It can be assumed that they are exposed to plastics and associated chemicals throughout their lifecycle, since coastal areas are among the most polluted marine habitats and close to the main entry points for plastics to the ocean (Manzo and Schiavo, 2022;Thushari and Senevirathna, 2020). Early life stages of sea urchins are most sensitive to stress (Pandori and Sorte, 2019;Przeslawski et al., 2015), while being decisive for population dynamics (Thorson, 1966). Recently, a review concluded that the main impact on sea urchins stems from released additives and sorbed chemicals rather than MP particles themselves (Manzo and Schiavo, 2022). The impact of plastic leachates on sea urchins has been studied for several materials Cormier et al., 2021;Martínez-Gómez et al., 2017;Nobre et al., 2015;Oliviero et al., 2019;Piccardo et al., 2021;Rendell-Bhatti et al., 2021;Uribe-Echeverría and Beiras, 2022), but not for tire particles.
This study aims to investigate the effects of tire particle leachates on early developmental stages of three keystone sea urchin species (Paracentrotus lividus, Arbacia lixula and Diadema africanum). These species are dominant in coastal areas of the Canary Islands and other Macaronesian islands, and they play important ecological roles, such as grazing control of benthic algae. This can strongly affect the benthic community structure of coastal habitats since the abundance of sea urchins determines the shift between the two major habitat states: erect algal forests and coralline barrens (Agnetta et al., 2013;García-Sanz et al., 2014;Gizzi et al., 2020;Wangensteen et al., 2011). Although co-occurring, the three species differ in their traits and ecological role. While P. lividus and A. lixula inhabit the intertidal and subtidal, D. africanum only inhabits the subtidal and generally occurs at greater depths (Tuya et al., 2007). P. lividus feeds more on flashy algae, while A. lixula with its superior Aristotle's lantern (thus greater grazing ability) feeds on encrusting coralline algae as well as invertebrates (Agnetta et al., 2013). D. africanum feeds both on macroalgae and encrusting algae (Sangil et al., 2014;Tuya et al., 2001). Of the three studied species, D. africanum is described as the dominant herbivore, able to generate big barrens with relatively low diversity (García-Sanz et al., 2014; Hernández et al., 2008). P. lividus can easily move into erect algal forests and create barren patches, which are then colonized by encrusting algae and cryptic fauna. Since this species is often characterized by cryptic behavior (meaning that it hides in crevices and stays within small patches), and more gentle grazing, it is weaker in maintaining the barren state. In contrast, A. lixula, less capable in creating barren patches, maintains these open areas through its more intense grazing and movement (Agnetta et al., 2013;Bonaviri et al., 2011). In the Canary Islands, D. africanum was for many decades the most abundant sea urchin species, creating extensive barrens, named locally "blanquizales" (white barren grounds). However, after two mass mortality events in 2009/2010 and 2018 its abundance decreased by 94-98% and many areas have shifted back to algal forests (Sangil and Hernández, 2022).
The overall hypothesis of this study was that the early developmental stages of sea urchins are negatively affected by tire leachates. Specifically, our research questions were: 1) Do tire particle leachates impair the growth, morphological development and survival of early life stages in sea urchins? 2) Does the sensitivity towards tire particle leachates differ between the three dominant sea urchin species of the Macaronesian islands? 3) Does the activity of selected biomarkers (the electron transport system (ETS) and glutathione S transferase (GST)) change under exposure to tire particle leachates?

Animal collection and spawning
Adult sea urchins (P. lividus, A. lixula, D. africanum) were collected by snorkeling in Arinaga, Gran Canaria, on the day of a given exposure experiment. The animals were transported to the laboratory in cool boxes with in situ seawater. Spawning was triggered by injecting 1 mL of a 0.55 M potassium chloride solution through the mouth. Immediately after, individuals were placed upside down on beakers filled with filtered seawater (FSW). All FSW used in this study had a salinity of 35 psu, 20 • C, pH 8.01 and an oxygen saturation of >90%. In case of a spawning male, strands of released sperm were immediately collected from the water column with a Pasteur pipette and transferred to a glass vial on ice. In case of a spawning female, oocytes were allowed to deposit in the beaker and once spawning diminished, they were transferred to a new beaker with FSW. The active swimming of sperm was checked under a microscope (DMi1 Leica) and the quality of the oocytes (evenness of size and spherical shape) was evaluated using a stereo microscope (Leica S APO). For fertilization, the oocytes of one to three females were mixed to create an oocyte suspension of roughly 2000 mL − 1 . Diluted sperm of up to three males was added to the suspension and fertilization success was confirmed by observing the development of a fertilization membrane around >90% of the eggs.

Preparation of tire particle leachates
The leachates were obtained as described in Almeda et al. (2023) with minor modifications in the micronization process (Page et al., 2022). Briefly, micronized tire particles (<250 μm, new Imperial 145/70-13 71T-Snowdragon HP tires) were kept frozen at − 20 • C until use. A suspension of 1 g L − 1 was prepared with FSW in a 500 mL glass bottle with a screw cap with a polytetrafluoroethylene protected seal. The bottle was completely filled and closed air-free. It was placed on a roller (LLG-UniROLLER 6 easy) at 15 rpm in an incubator at 20 • C in darkness. After 72 h, the suspension was filtered over a 0.7 μm glass fiber filter to remove the particles. The filtered liquid constituted the leachate stock solution (i.e., 100%) of which the following dilutions were prepared: 75%, 50%, 25%, 12.5% and 6.25% (corresponding to 0.75, 0.5, 0.25, 0.125 and 0.0625 g L − 1 ). Throughout the paper, we refer to these dilutions in their equivalent mass concentrations. The pH of the leachate stock, as well as the FSW, was checked in triplicates to ensure that the pH did not differ between treatments (pH 7.87 ± 0.1 in leachates, pH 7.89 ± 0.1 in FSW, Crison GLP21).

Chemical analysis of leachates
Undiluted leachates were analyzed for 52 organic compounds (Table S1) and 50 elements (Table S3). Details on the used reagents are available in the supplementary material.

Organic compounds analysis
After the preparation of the leachates, subsamples of the leachate stock solution (1 g L − 1 ) for organic analysis were immediately frozen at − 20 • C. For the analysis, leachate samples were thawed and placed on an orbital shaker for 2 h. An aliquot of 50 mL was transferred into another vessel, and 10 μL of the procedural internal standard (P-IS) solution (containing acenaphthalene-d10, chrysene-d12, phenanthrene-d10, and PCB 200) was added. The solution was left in darkness for 1 h. All samples were extracted and analyzed in triplicate. Then, they were submitted to solid-phase extraction using 200 mg (3 mL) Chromabond® C18ec columns (Macherey-Nagel, Germany) mounted in a vacuum manifold (Waters Corporation, USA). Before applying the samples, the columns were cleaned and conditioned with 2 mL of methanol and 2 mL of water/n-propanol (85:15, v/v). Next, 50 mL of the leachates were passed through the cartridge under vacuum at a flow rate of 1.5 mL min − 1 . The column was washed with 1 mL of the water/n-propanol solution, and dried under high vacuum for 30 min. Finally, the compounds were eluted from the column with 1 mL of dichloromethane.
Quality controls and blanks were prepared in autoclaved seawater using the same methodology. They were spiked 1 h before the extraction with the mix standard solution and/or with the P-IS solution, respectively. A 10-point calibration curve in dichloromethane ranging between 0.195 and 100 ng mL − 1 was used for quantification of the organic compounds.
The analysis was performed using a 7890B gas chromatography (GC) system coupled to a Triple Quad 7010 mass spectrometer (MS) (Agilent Technologies, USA). All GC parameters have been described previously (Camacho et al., 2019) and retention times, precursor, fragment ions, and collision energies for each compound are listed in the supplementary material.

Elemental analysis
An aliquot of 40 mL of leachates from the stock solution (1 g L − 1 ) was transferred into a 50 mL falcon tube, vortexed, and placed in an ultrasonic bath for 30 min. Once sonicated, the samples were centrifuged at 2500 rpm for 3 min. A direct dilution of the sample was made at 1:10, using a solution of nitric acid (HNO 3 ) (2%). For this purpose, 1800 μL of 2% HNO 3 and an aliquot of 200 μL of the sample were added to the vial and shaken.
Two complementary calibration curves covering the range 0.03-300 ng mL − 1 were used. The elemental analysis was performed using an Agilent 7900 inductively coupled plasma mass spectrometer (ICP-MS) equipped with standard nickel cones and a crossflow nebulizer with a make-up gas port (×400 nebulizer, Savillex Corporation, USA) for all measurements. Daily, a tuning solution consisting of a mixture of cesium, cobalt, lithium, magnesium, thallium and yttrium (Agilent Technologies) was used to optimize the ICP-MS instrument. The working conditions were assessed by considering the sensitivity of the six isotopes in counts per second, the associated relative standard deviation, oxide ratio, and double ion charges. The entire procedure has been described in detail previously (Rodríguez-Hernández et al., 2019).

Exposure experiment
A 72 h exposure of developing embryo/larvae to the six leachate dilutions plus a control (FSW) was conducted. Small (68 mL) glass bottles were filled with 30 mL of the respective dilution or FSW and fertilized oocytes were added to reach a density of 30 mL − 1 . All treatments were done in triplicate. The bottles were loosely covered with aluminum foil and placed in an incubator at 20 • C in darkness.
For P. lividus, additional exposure bottles of four treatments (25%, 12.5%, 6.25% and control) were prepared for enzymatic assays (see below). In this case, the volume per bottle was 200 mL with a density of 50 fertilized oocytes per mL to obtain enough biomass for later analysis. Except for this difference, the bottles were prepared and treated in the same way as the small ones.
After 72 h, we quantified the percentage of mortality and abnormal development (developmental arrest, strong skeletal deformations, missing arms; see Rendell-Bhatti et al., 2021). In a subsample of 1 mL, the total number of larvae, and the number of dead and abnormally developed larvae was determined under a stereo microscope. In parallel, 12 mL of each bottle were transferred to a falcon tube with the addition of four drops of 10% formalin. After the larvae had settled, they were transferred to a glass slide and images were taken on a microscope with a connected camera (Leica Flexacam C3). For each replicate, the length (from the apical end to the connection between the anterolateral arms) of 20 randomly selected larvae was measured using the software ImageJ.
The total content of the 200 mL bottles was filtered on a 50 μm plankton mesh and the larvae were transferred to a pre-weighed 2 mL Eppendorf tube using a fine spatula. Immediately, the wet weight was recorded and the tube was frozen at − 80 • C.

Biochemical analysis of GST and ETS activity
Proteins were extracted by homogenizing the whole filtrate of larvae in phosphate buffer (0.1 M, pH 7.8) with 0.2% Triton X100 using an ultrasonic tissue homogenizer. The oxidative stress marker glutathione S transferase (GST) was assayed in protein extracts at 340 nm on a BMG-FLUOstar omega microplate reader following the microplate-adapted procedure (Frasco and Guilhermino, 2002) from the original work of Habig et al. (1974). The activity of the electron transport system (ETS) was measured in larvae homogenates following the production of INT (2-(p-iodophenyl)-3-(p-nitrophenyl)-5-phenyl tetrazolium chloride)formazan at 490 nm using a Cary series UV-VIS spectrophotometer, according to Owens and King (1975) and the modified procedure of Gómez et al. (1996).
Enzyme activity was calculated with a molar extinction coefficient of 9.6 mM − 1 cm − 1 and 14.9 mM − 1 cm − 1 for GST and ETS activity, respectively. The activity of ETS was then stoichiometrically related to oxygen consumption . The Pierce BCA protein assay kit (ThermoFisher Scientific) was used to measure the protein concentration of larvae homogenates, and results were expressed as specific activity (μmol mL − 1 mg of protein − 1 for GST and μL O 2 h − 1 mg of protein − 1 for ETS activity).

Data analysis
All statistical analyses were performed using the software R (version 3.6.3, R Core Team, 2022). Analysis of variance (ANOVA) was used to test for differences between means of treatments in the size of larvae, enzymatic activity, and mortality (for P. lividus and A. lixula). The assumptions of normality and homogeneity of variances were tested with the Shapiro-Wilks and Fligner-Kileen test, respectively. As a posthoc test Tukeys HSD was used. Dose-response curves were modeled with a four-parameter log-logistic function (drc R package, Ritz et al., 2015) with their associated 95% interval calculated with Delta's method, using abnormal development data of P. lividus and A. lixula and mortality data of D. africanum. The absolute half maximal effective dilution (ED 50, P. lividus and A. lixula) and absolute half maximal lethal dilution (LD 50, D. africanum) were calculated using respective models. Based on this, the estimated LD 50 and ED 50 (%) are expressed in their equivalent mass concentrations in g of tire particles L − 1 (LC 50 and EC 50 ) to facilitate the comparison with other toxicity studies and the evaluation of the risk considering the environmental concentrations of TWPs. LC 50 and EC 50 values were also used to determine and rank species sensitivity. Lethal effects (reflected in a LC 50 ) were ranked as more severe, i.e. implying a higher sensitivity, than sub-lethal effects (reflected in an EC 50 ).

Tire particle leachates affect normal development and survival of sea urchin larvae
We found clear concentration-dependent effects of tire particle leachates on all species. In the controls, all individuals were alive at the end of the exposure and very few showed signs of abnormal development (2.2% of P. lividus, 3.8% of A. lixula and 4.5% of D. africanum, respectively) (Fig. 1). In P. lividus, the fraction of normally developed larvae steadily decreased with increasing leachate concentration with a minimal percentage of 26.7% normally developed larvae at 0.25 g L − 1 . From a leachate concentration of 0.5 g L − 1 on, no normally developed larvae were found. At the same time, mortality was observed from this concentration and up, with a maximum level of 36.9% mortality in the undiluted leachate (1 g L − 1 ) (Fig. 1A). In contrast to P. lividus, both A. lixula and D. africanum showed high levels of normal development (79-96% and 89-96%, respectively) until a leachate concentration of 0.25 g L − 1 . However, from 0.25 to 0.5 g L − 1 both species displayed a marked increase in abnormal development and mortality. For A. lixula, the majority of larvae was abnormally developed from a leachate concentration of 0.5 g L − 1 on and there was a marked increase in dead larvae (14.3% at 0.5 g L − 1 to 24.7% at 1 g L − 1 ) (Fig. 1B). Mortality of larvae started appearing at lower leachate concentrations for D. africanum (from 0.06 g L − 1 ) but increased dramatically from 6.7 to 63.7% dead larvae at 0.25 and 0.5 g L − 1 , respectively. In the two highest concentrations, all larvae were dead at the end of the exposure (Fig. 1C).
The percentage of abnormally developed larvae was used to plot dose-response curves and calculate EC 50 values for P. lividus and A. lixula (Fig. 2). After 72 h, normally developed larvae had reached a four-arm pluteus stage (Fig. 2B). Malformations were often very pronounced, including malformed or missing arms, which indicates impaired skeleton development. Similar phenotypes were observed when P. lividus was exposed to leachates of PVC, beached nurdles and biobeads (Rendell-Bhatti et al., 2021). The EC 50 values for P. lividus and A. lixula were 0.158 g L − 1 and 0.345 g L − 1 , respectively, indicating a higher sensitivity of the former.
The effects of tire leachates on early development of marine invertebrates have so far only been studied in bivalves. Capolupo et al. (2020) observed abnormal development in the Mediterranean mussel M. galloprovincialis after 48 h with an EC 50 of 1.8 g L − 1 . After 36 h exposure, the embryonic development of oysters (C. gigas) was significantly impaired at a concentration of 10 g L − 1 (Tallec et al., 2022b). These effect concentrations are one to two orders of magnitude higher than our findings, which suggests that sea urchin larvae are more sensitive to tire particle leachates than bivalve larvae. Differences can, however, also be related to the individual composition and concentrations of chemicals in the leachates. This depends on the used tire material and may also be influenced by the used protocol for preparing the leachates (e.g., tire particle size, particle concentration, incubation time). Thus, standardizing the protocols in leachate studies as proposed by Almeda et al. (2023) is an important step to increase comparability.
For D. africanum, effects on development were overridden by mortality at higher leachate concentrations. Since dead larvae had usually started degrading by the end of the experiment, it was not possible to determine their phenotype (normally vs. abnormally developed) with certainty. Instead of an EC 50 , we therefore calculated a LC 50 , which was 0.456 g L − 1 (Fig. 2D). As mentioned above, some degree of mortality was also observed for the other two species, but nowhere near those of D. africanum. Since lethal effects are more severe than sub-lethal effects, D. africanum was the most sensitive of the tested species. The observed differences between species clearly demonstrate that responses are highly species-specific and it could indicate differing stress response mechanisms or uptake of chemicals among species.
Effects of tire particle leachates on survival have also been found for other marine zooplankton. Reported LC 50 values are 47.5 g L − 1 for larvae of M. galloprovincialis (Capolupo et al., 2020), 5.34 g L − 1 for the copepod Tigriopus japonicas (Yang et al., 2022), 35 g L − 1 and <5 g L − 1 for the copepods Calanus and Acartia respectively (Halsband et al., 2020), and 0.6 g L − 1 for the rotifer Brachionus plicatilis (Shin et al., 2022). Similar to our EC 50 findings, the obtained LC 50 value is considerably lower than most reported ones in the literature. Although differences in LC 50 /EC 50 can be due to differences in the methodology as described above, our results indicate that early life stages of sea urchins are very sensitive to tire particle leachates compared to other zooplankton. This is in accordance with sea urchin embryos and larvae generally being acknowledged as highly sensitive to pollutants and therefore being used as standard test organisms in ecotoxicology (Environment Canada, 2011; Morroni et al., 2016). However, the most sensitive species will depend on the pollutant and it is important to study effects in a range of organisms (Bellas et al., 2005). The importance of multi-species studies has also been demonstrated in the work with tire leachates and fish. For instance, McIntyre et al. (2021) found acute toxicity of TWP leachates to coho (Oncorhynchus kisutch) but not closely related chum salmon (O. keta). Also, the tire-derived chemical 6PPD-quinone resulted in mortality of some fish species, but not others Brinkmann et al., 2022;Foldvik et al., 2022). In line with our observations, this highlights the high species-specificity towards tire leachates and

Fig. 1. Percentage of larvae of Paracentrotus lividus (A), Arbacia lixula (B) and
Diadema africanum (C) that were normally developed, abnormally developed or dead at the end of 72 h exposure to a range of dilutions of tire particle leachates (100% = 1 g L − 1 ). associated chemicals.

Impact of tire particle leachates on larval growth
The larvae of all species exhibited a concentration-dependent decrease in body size (Fig. 3). This was most pronounced for P. lividus, for which the reduction in comparison to the control was up to 49% (Fig. 3A), while this was 44% for A. lixula (Figs. 3B) and 35% for D. africanum (at 0.75 g L − 1 since larvae in the undiluted leachate (1 g L − 1 ) had fully degraded by the end of the experiment, Fig. 3C). For both A. lixula and D. africanum, larvae were significantly smaller than the control from a concentration of 0.5 g L − 1 on. These are the same leachate concentrations at which marked changes in abnormal development and survival were observed ( Fig. 1B and C). In contrast, larval size of P. lividus already decreased significantly by 10% at 0.125 g L − 1 . The high sensitivity of P. lividus here is in line with the findings of abnormal development.
A decrease in larval size of P. lividus was also found with leachates of environmental MPs, though from a concentration of 3.3 g L − 1 (Cormier et al., 2021). Furthermore, leachates of PVC plastic toys reduced larval  growth, with EC 50 values between 3.3 and 14.3 g L − 1 (Oliviero et al., 2019). In juvenile oysters it was found that exposure to TWP leachates reduced their scope for growth by 57% at a concentration as low as 1 mg L − 1 due to reductions in clearance and respiration rates (Tallec et al., 2022a). Since we performed this experiment in the absence of food, differential feeding cannot be a cause for the observed decrease in larval growth.

Enzymatic responses to tire particle leachates
Glutathione S transferase (GST) activity tended to increase in the leachate-exposed larvae of P. lividus compared to the control (Fig. 4A), reaching approximately twice the activity at a leachate concentration of 0.25 g L − 1 . This well-known enzyme involved in detoxification is often used as a biomarker of environmental contamination as it facilitates the elimination of xenobiotic compounds through catalyzing their conjugation to the antioxidant glutathione (Cunha et al., 2005). Thus, the observed trend could indicate that a detoxification response involving GST was induced by tire particle leachates, although the increase was not statistically significant (p = 0.07). Other studies report that the GST activity of P. lividus larvae can be induced, inhibited, or unaffected depending on the nature and the intensity of the chemical stress (Bellas et al., 2022;el Idrissi et al., 2022). Taking together our observations and reports in the literature, GST can serve as a sub-lethal endpoint in exposure studies with pollutants and may give insights into molecular processes. However, more work is needed to confirm the response of GST to tire particle leachates and other plastic-associated chemicals in echinoderm larvae.
The activity of the electron transport system (ETS) (index of respiratory activity) ranged from 0.4 to 0.9 μmol O 2 h − 1 mg of protein − 1 in larvae of P. lividus, with no clear trend in relation to exposure condition (Fig. 4B). No significant effect of the leachate was found. Values were in accordance with those reported by Tomšić et al. (2011) in early stages of P. lividus. The absence of a response in this parameter could indicate that the respiration rate remained constant while maintenance costs might have increased due to detoxification processes (e.g., GST induction). Therefore, larvae may either have reduced energy invested in maturation and structural growth to favor maintenance cost (Jager et al., 2016), or increased the energy produced by anaerobic means. The former could partly explain the observed decrease in larval growth and impairments in development.

Potential drivers of chemical toxicity
Leachates are highly complex mixtures of chemicals (Müller et al., 2022). Thus, it is very challenging to determine which compound(s) cause the observed effects. Most likely, the effects are a result of the presence and interactions between multiple chemicals, which hampers interpretation. For instance, Yang et al. (2022) observed an antagonistic effect of zinc and benzothiazole (a common organic chemical in tires) on copepods. Furthermore, the results from our chemical analysis cannot be generalized since the chemical composition of the tire material and its leachate depends on many factors, including tire age, brand, and the lixiviation protocol.
The chemical analysis of our leachates revealed elevated concentrations of seven organic chemicals and eight metals (Table 1). Among the organics, pyrene (0.48 μg L − 1 ), tris (2-chloroisopropyl)phosphate (0.43 μg L − 1 , TCPP) and naphthalene (0.22 μg L − 1 ) exhibited the highest concentrations. In a previous study, pyrene and naphthalene were found to have a low toxicity to larvae of P. lividus with an EC 10 and EC 50 value of 68.8 μg L − 1 and 4.7 mg L − 1 , respectively (Bellas et al., 2008). These are two and four orders of magnitude above our detected concentrations. Our observed effects are therefore unlikely to be caused by pyrene or naphthalene, at least for this species. TCPP is a common additive (flame retardant) in plastics and considered a chemical of emerging concern. Very little toxicity data on marine invertebrates is available to date. In adults of P. lividus, TCPP was found to affect the gonad development at 1 μg L − 1 and above (Campoy-López et al., 2020). This is more than double the concentration in our tire particle leachate. However, early life stages are often more sensitive than adults and the toxicity of TCPP to sea urchin embryos and larvae remains to be studied.
The analysis of metals revealed elevated concentrations of barium and zinc in high concentrations (88.43 and 43.98 μg L − 1 , respectively), nickel, manganese, and iron in low concentrations (2-8 μg L − 1 ), chromium, lead, and copper in very low concentrations (<1 μg L − 1 ). The elevated concentration of zinc is in accordance with other studies that found zinc as the most prominent metal in tire particle leachates (Capolupo et al., 2020;Halsband et al., 2020;Tallec et al., 2022b). This is not surprising as zinc is a common additive in tires, serving as a vulcanization activator (Wik and Dave, 2009). Zinc was identified as the main toxicant of TWP leachates for the copepod T. japonicus (Yang et al., 2022). It is also known to be toxic to sea urchin larvae, which is likely due to a disruption of calcium homeostasis (Tellis et al., 2014). Radenac treatment. Novelli et al. (2003) reported an EC 50 of 49 μg L − 1 for the 72 h development of P. lividus. Zinc is therefore very likely to explain some of our observations. Reported EC 50 values for other sea urchin larvae are 96.9 μg L − 1 (Phillips et al., 2003) and 151 μg L − 1 (Nadella et al., 2013) for Strongylocentrotus purpuratus, and 27.7 μg L − 1 for Evechinus chloroticus (Rouchon and Phillips, 2017). These results again highlight the species-specific sensitivity of sea urchin larvae. The effect of zinc has also been found to differ in mixtures with other metals, showing that both antagonistic and synergistic effects are possible (Nogueira et al., 2021;Phillips et al., 2003;Rouchon and Phillips, 2017). The binary mixture of zinc and nickel for instance strongly affected Ca 2+ influx in S. purpuratus (Nogueira et al., 2021). The same study also reported strong effects of nickel and the zinc-nickel mixture on larval development. This was observed at a nickel concentration of 5 μg L − 1 , which roughly corresponds to our 50% (0.5 g L − 1 ) treatment. Therefore, nickel as well is a strong candidate to have caused some of the observed effects. Barium, the most highly concentrated metal in our leachates, is known to be deposited in the skeleton (Moore, 1991) but few studies have investigated the toxicity on marine invertebrate larvae. Spangenberg and Cherr (1996) observed abnormal larval development of Mytilus californianus at concentrations between 200 and 900 μg L − 1 , more than twice our concentration. Lethal effects on embryos of the yellow crab (Cancer anthonyi) were observed at 100 mg L − 1 (Macdonald et al., 1988). Although barium therefore seems unlikely to be a driver for the observed toxicity, its effect on sea urchin larvae remains to be studied.
As the tire particle leachates are a complex mixture of chemicals, it is almost certain that more compounds were present in our leachates than those we detected with target analysis, resulting in potentially unknown effect mechanisms. As a next step we therefore suggest using a method coupling chemical fractionation associated with bio-tests, such as Effect-Directed Analysis, to identify chemicals that individually or associated with others are responsible for the observed toxicity (Burgess et al., 2013).

Ecological implications
Planktonic larvae are decisive for the population dynamics of benthic invertebrate populations. They already face high levels of mortality due to predation and transport to unsuitable habitats (Thorson, 1950). If their development is additionally impaired by chemical pollution, this will decrease the chance of successful recruitment. The three studied species of echinoids play key roles in the dynamics of benthic coastal rocky ecosystems as major grazers of benthic algae. Negative impacts on their abundance through a decrease in recruitment would directly affect the degree of herbivory in those coastal benthic ecosystems, potentially changing the dynamics of algal forests and coralline barrens (Agnetta et al., 2013). This would have consequences for the whole benthic community which differs between these two states. At low concentrations, P. lividus was the most affected species as it was already significantly impaired at a leachate concentration of 0.125 g L − 1 . A. lixula and D. africanum were not affected at this level. A reduction of P. lividus could result in fewer barren patches within algal forests, decreasing heterogeneity and the habitat for associated encrusting and cryptic organisms. At higher concentrations of leachates (from 0.5 g L − 1 ), D. africanum was most susceptible, showing high levels of larval mortality. Such a 'high pollution scenario' could threaten the occurrence of this species locally, especially after the recent mass mortalities (Sangil and Hernández, 2022). However, these leachate concentrations are unlikely in the water column based on current knowledge.
The cause for the observed difference in sensitivity between species is difficult to determine. Although not quantified, we observed a marked difference in egg size and color. While those of P. lividus and A. lixula only differed in color (orange vs. purple), D. africanum had substantially smaller and paler eggs, even though similar egg sizes for all three species have been reported (Byrne, 1990;George et al., 1990;Hernández et al., 2020). Intraspecific variation can be related to genetic differences or environmental influences (Moran and McAlister, 2009). Since egg size/volume is related to energy content (McAlister and Moran, 2012;Moran and McAlister, 2009), the larvae of D. africanum may have had lower energy reserves, potentially explaining the stronger response in mortality. The coloration of eggs is related to different pigments, which can have antioxidant properties such as astaxanthin in A. lixula (Galasso et al., 2018;Limatola et al., 2021). However, the role of pigments in the stress tolerance of sea urchin embryos and larvae is unclear. Different sensitivity of species that live in the same habitat could ultimately lead to changes in the community structure in polluted areas.
Field data on the concentration of tire particles in coastal environments is still very scarce, making it difficult to evaluate their ecological risk. Wik and Dave (2009) estimated a predicted environmental TWP concentration in surface waters of up to 56 mg L − 1 . Based on these data, the range of equivalent concentrations used in our exposure concentrations (0.06-1.0 g L − 1 ) are environmentally relevant, particularly after runoffs events in coastal shallow waters. Concentrations of TWPs of 125 mg L − 1 and above are currently unlikely in marine waters. However, it is important to identify local point sources. For instance, coastal areas with heavy traffic and/or the release of untreated storm runoff water can be major entry sources for TWPs. Furthermore, the growing evidence for detrimental effects of leachates from tires questions their intentional deployment, for instance, to build artificial reefs and to be used in ports (Collins, 2021). As with other MPs, TWPs are expected to sink to the seafloor reaching higher concentrations than in the water column.

Table 1
Chemical composition of the undiluted tire particle leachates (100%, 1 g L − 1 ) and of the filtered seawater. All organic compounds and metals are shown for which concentrations in the leachate exceeded those of the blank. All values in μg L − 1 . a, b, c indicate replicates. There, they may continue leaching their toxic chemicals into the water. Leads and Weinstein (2019) found up to 2.3 mg TWPs kg − 1 in subtidal sediments, 0.29 mg m − 2 in intertidal sediments and 0.01 mg L − 1 in the sea surface microlayer of the Charleston Habor Estuary, USA. Other studies reported TWP mass concentrations of up to 14.5 g L − 1 in snow meltwater (Rødland et al., 2022), 150 mg g − 1 in gully pot sediment (Mengistu et al., 2021), 155 mg g − 1 in river sediments (Wik and Dave, 2009) and 280 mg L − 1 in stormwater runoff (2-125 μm TWPs, Järlskog et al., 2022). There is still a high level of uncertainty concerning the impact of TWPs on marine ecosystems. This is complicated by the fact that chemical composition differs between brands and depends on the use time and weathering state of the material (Halsband et al., 2020;Kim et al., 2023). As mentioned above, this hampers the comparability between leachate studies and making predictions for natural ecosystems based on their results. Furthermore, leached chemicals can be transformed in the environment, resulting in enhanced toxicity . Therefore, data on the concentration, persistence and fate of TWPs as well as their associated chemicals in the environment is urgently needed to better evaluate the potential impacts of these pollutants on aquatic biota.

Conclusions
Early life stages of three sea urchin species were negatively affected by acute exposure to tire particle leachates, resulting in reduced growth, abnormal development, and mortality. The sensitivity towards the leachates differed among species. Overall, D. africanum was the most sensitive species with high levels of mortality (LC 50 0.46 g L − 1 ), followed by P. lividus (EC 50 0.16 g L − 1 ) and A. lixula (EC 50 0.35 g L − 1 ) based on the abnormal development of the latter two. There was however a threshold for D. africanum at a leachate concentration of around 0.5 g L − 1 , below which this species did not show effects. At these low concentrations, P. lividus was most strongly impacted. This reflects the importance of conducting multi-species studies to determine the toxic effects of contaminants on natural marine communities. The three species play different roles in shaping coastal benthic habitats through their grazing activity and changes in their fitness or abundance can have profound consequences for the entire benthic community. The responses were observed at concentrations that are slightly above currently measured and predicted levels. However, data on TWPs and their leached chemicals in the marine environment is still extremely limited, and more research is urgently needed to evaluate the impacts of TWP pollution on marine ecosystems.

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability
Data will be made available on request.