Commonly co-occurring stressors in industrialized estuaries disrupt larval development in blue mussels

1 Introduction

Estuaries are increasingly exposed to the effects of industrialization, which lead to multiple, interacting stressors that threaten to undermine ecological functioning (Halpern et al., 2008). Indeed, long–term, intense human activity has led to the degradation of estuarine environments, leaving associated sediments contaminated with so-called “legacy pollutants” including heavy metals, polyaromatic hydrocarbons (PAHs) and polychlorinated biphenyls (PCBs; Rodgers et al., 2020; Yates et al., 2007; Sheahan et al., 2001) that continue to leach into the water column long after primary pollution discharge ceases (Bardos et al., 2020). Furthermore, the proliferation of emerging contaminants (ECs)—such as pharmaceuticals, pesticides and synthetic materials including per–/poly–fluoroalkyl substances (PFAS) and microplastics (Sultan et al., 2024)—add further chemical pressures to estuarine environments, the ecological impacts of which are often poorly understood (Drakvik et al., 2020; Letsinger et al., 2019). Respective contaminants almost never appear in isolation (Moore et al., 2021); instead, they form complex mixtures capable of exerting potentiating effects on aquatic organisms, even when individual components occur at concentrations considered “safe” (Al-Shaeri et al., 2013; Cedergreen, 2014). Yet mixture risk assessments remain incomplete and do not reflect the complexity of pollution within European brackish ecosystems (Posthuma et al., 2019).

Climate change is expected to intensify pollution effects, with growing consensus that rising temperatures and salinity decreases, projected for many northern European estuaries, will push organisms toward their physiological limits (IPCC, 2023; Dye et al., 2020) while simultaneously modulating contaminant bioavailability, speciation and toxicity (Noyes et al., 2009; Kazmi et al., 2022). Increased temperature can enhance chemical uptake and toxicity, influence diffusion rates, and ultimately shape the joint toxicities of multiple chemicals (reviewed in: Kazmi et al., 2022). Salinity has a direct impact on bioavailability of contaminants, with reduced salinity causing an increase in metal accumulation and toxicity in numerous aquatic organisms and has also been shown to alter solubility of PAHs in the water column, affecting exposure pathways (reviewed in: Schiedek et al., 2007). Combined osmotic stress and chemical exposure has been shown to induce abnormalities in embryogenesis across multiple marine species, for example, increased toxicity of metolachlor and copper to embryo–larval development in Pacific Oysters (Crassostrea gigas; Gamain et al., 2016) and enhanced toxicity of heavy metals (copper, zinc and cadmium) to bivalve larvae including M. edulis (McLusky et al., 1986; Zhadan et al., 1992). These interactions indicate that climate–driven abiotic change may amplify the biological impacts of pollution, creating stressor combinations more harmful to estuarine organisms than individual stressors alone (Holmstrup et al., 2010; Noyes et al., 2009). Notwithstanding, policymakers often treat chemical pollution as distinct from non-chemical stressors (Posthuma et al., 2019), creating a persistent divide in research on how multiple stressors and chemical mixtures shape biotic responses, despite these pressures co-occurring spatially and temporally with well-documented joint effects (Schäfer et al., 2023).

Early life stages of marine organisms are typically the most sensitive to fluctuations in environmental conditions and chemical stressors, with vulnerability generally decreasing as organisms mature (Gamain et al., 2016; Kruft Welton et al., 2024). For instance, bivalve embryos and larvae can be two to three orders of magnitude more sensitive to heavy metals than adults (Saidov and Kosevich, 2019). Multiple stressor interactions are therefore likely to disproportionately affect larvae, particularly in species with pelagic larval stages that are directly exposed to interacting stressors in the water column. Yet, most multiple-stressor research has focused on adult life stages (Przeslawski et al., 2015), despite the likelihood that stressor interactions will act through different mechanisms in embryos and larvae. Developmental disruption that occurs during early life stages can create recruitment bottlenecks that ultimately drive population declines (Pechenik, 1999; Gosselin and Qian, 1997). Indeed, recent declines in blue mussel (Mytilus edulis) populations in the eastern and south-western Scottish estuaries (Burrows et al., 2025) have raised concerns that interacting stressors, rather than single drivers, may be undermining successful recruitment in more industrialized estuaries (Jaatinen et al., 2021; Lojuyo et al., 2022).

This study investigated how early developmental stages of M. edulis (fertilized egg to D–shape veliger stage) respond to an environmentally relevant pollutant mixture under temperature and salinity conditions projected for estuaries across the UK (IPCC, 2023; Dye et al., 2020). Based on the mechanisms outlined above, we hypothesized that climate-driven changes would interact with the pollutant mixture to produce deleterious effects on larval development. Specifically, we predicted that elevated temperature and reduced salinity, both individually and in combination, would exacerbate pollutant toxicity, reducing normal D-stage veliger development and increasing the frequency of shell malformations.

2 Materials and methods

Fertilized eggs harvested from M. edulis were reared under normal conditions (‘Control', i.e., 15 °C and 35 ppt) or exposed to elevated temperature (18 °C) or reduced salinity (25 ppt), both alone or in combination for 48 h. The two temperatures and salinity levels were selected based on average conditions in the Firth of Forth, UK (~13 °C−15 °C and ~ 34 ppt in the summer; Marine Scotland, 2025), projected climate change (2 °C−4 °C warming by the end of the century–projected by the IPCC under a high emissions scenario [SSP5–8.5]; IPCC, 2023, and overall decrease in salinity; Dye et al., 2020). Larvae were exposed to each of these conditions (n = 4: 1. Control; 2. 18 °C + 35 ppt; 3. 15 °C + 25 ppt; 4. 18 °C + 25 ppt) and to six levels of an environmentally relevant pollutant mixture (Trace, Low, Environmental, Moderate, High and Extreme, see Tables 1 and 2 for concentrations). The pollutant mixture was designed to be both “environmentally relevant,” and, to form a concentration-response gradient using higher and lower exposure concentrations.

Table 1. Proportion of components within the metal mixture, concentrations of each component within each exposure level (Trace-Extreme) and EC50 values for normal larval development from the literature (Martin et al., 1981; Saidov and Kosevich, 2019).

Table 2. Proportion of components within the organic mixture, concentrations of each component within each exposure level (Trace-Extreme) and EC50 values for normal larval development in Mytilus spp. from the literature (Balbi et al., 2018; Franzellitti et al., 2019; Hayman et al., 2021; Russo et al., 2022).

2.1 Pollutant mixtures

Pollutant mixtures were designed using information collated from secondary data sources (Tables 1 and 2) to capture realistic exposure ranges. Focal pollutants were selected based on a high detection frequency in anthropogenically impacted estuarine waters in the UK, using methods outlined by Moore et al. (2021), with proportions of chemicals within the mixture selected using published median concentrations of emerging and legacy pollutants (see Supplementary Table 1). This composite approach provides an environmentally realistic representation of typical contaminant levels in industrialized estuaries while allowing controlled experimental manipulation of mixture composition. The Environmental concentration reflects a generalized contaminant scenario, rather than the exact chemical profile of a single location.

Components were purchased from Sigma Aldrich (Poole, UK) and LGC Standards and are detailed in the Supplementary material. Individual stock solutions were made for each component of the pollutant mixture in compatible solvents: acetone (9 components—organic compounds), ethanol (7 components—organic compounds), and Milli-Q water (10 components—metals). Individual stock solutions with the same solvent were mixed together then the subsequent mixtures were diluted 1:100. Serial dilutions to the desired concentrations were used to prepare stocks for the lower exposure concentrations. The organic mixtures (ethanol and acetone) were then further diluted 1:100 in seawater so each exposure had equivalent solvent volume (0.05% solvent) when 25 μL of each mixture was added to a final volume of 2.5 mL per well.

2.2 Mussels and gamete collection

Sexually mature, rope-grown blue mussels (Shetland, UK; Seaspray Mussels) were purchased from a local fishmonger in Edinburgh (UK) in June 2025 and transported live on ice to the laboratory. Mussels were immediately cleaned of epibionts and byssal threads were removed. Individuals were then placed in aerated natural seawater at pH 7.9–8.2, 33 ppt, 14 °C ± 1 °C to induce spawning by thermal shock (ASTM, 2004). When gamete release was observed, mussels were placed in individual containers filled with 250 mL of natural seawater until complete gamete emission (Balbi et al., 2018). Mussels were removed from the containers post spawning, and all eggs and sperm were sieved into 250 mL beakers (keeping eggs and sperm from different individuals separate) using 100 μm and 37 μm meshes, respectively, to remove any feces or debris (Balbi et al., 2018; Fabbri et al., 2014; Saurel et al., 2022).

Samples of eggs or sperm from each individual were examined under a light microscope in test chambers to determine egg quality (shape, size) and sperm motility (Balbi et al., 2018). Once checked, eggs and sperm from two individuals were paired and sperm was gradually added to the eggs at a ratio of 1:10 (ASTM, 2004). After 30 min, fertilization success (number of fertilized eggs divided by the number of total eggs multiplied by 100) was determined under a light microscope through the observation of polar body development and only eggs with a fecundation rate above 85% were deemed acceptable (Balbi et al., 2018). The egg-sperm pairing with the highest percentage fertilization success was then chosen for use in the subsequent 48-h bivalve toxicity test, to minimize genetic and gamete variability to improve consistency in the biomarker response (Damiens et al., 2004).

2.3 Toxicity tests

48-h bivalve toxicity tests took place in 24-well sterile, capped polystyrene plates (Costar, Corning Incorporate, NY, USA) using a certified procedure (ASTM, 2004). Fertilized eggs were added to the appropriate salinity natural seawater (35 or 25 ppt, adjusted with distilled water, aerated, pH 7.9–8.2) with approximately 30 eggs per well. There were five replicate wells for each exposure combination. Appropriate treatments (trace to extreme) were added to each well in a random matrix to achieve the desired pollutant concentration at a final volume of 2.5 mL. Additionally, as the components of the pollutant mixture were dissolved in either ethanol, acetone or Milli-Q^® water, a solvent control (final concentration of 0.01% ethanol, 0.01% acetone, and 1% Milli-Q^® water in each well) was performed simultaneously and compared against the control conditions to confirm these solvents did not significantly impair the development of Mytilus spp. across all abiotic treatments (see Supplementary Figure 1). Wells were then incubated at the appropriate experimental temperature (15 °C ± 1 °C or 18 °C ± 1 °C) for 48 h. Following incubation, larvae were fixed in buffered formalin (4%). Normal development was assessed by counting larvae using an inverted Olympus IX70 microscope (Olympus, Tokyo, Japan) at 40x magnification. Larvae were considered normal if they had developed a D-shape shell with a straight hinge (see Supplementary Figure 2F) within the 48-h period, and malformed if they had developed abnormalities (Fabbri et al., 2014) such as a convex hinge, protruded mantle, indented margin or incomplete shell (see Supplementary Figure 3) or remained in the earlier trochophore stage (see Supplementary Figure 2E). The proportions of normal, malformed and trochophore larvae were then calculated and used to assess the toxicity of the pollutant mixture under different climatic conditions.

2.4 Statistical analyses

All statistical analyses were conducted in R (v.4.3.2). All data are expressed as mean percentages ± standard deviation (SD). Binomial generalized linear models (GLMs) were fitted to the raw larval count data to assess salinity and temperature influence on probability of successful development (normal vs. abnormal), with quasibinomial correction applied for mild overdispersion (Pearson χ²/df = 1.21). ANOVAs on percentage data were run in parallel to facilitate interpretation of stressor interactions across all development endpoints (normal D-stage, malformed D-stage and trochophore), with Shapiro-Wilk tests and Levene's tests used to check for normality and homogeneity of variance, respectively. Post-hoc pairwise t-tests were used to compare differences between conditions within each pollutant-exposure level. Statistical significance was accepted at p < 0.05.

3 Results

3.1 Temperature and salinity stress response

Reduced salinity and elevated temperature significantly affected the probability of normal development (quasibinomial GLM: salinity, z = 4.775, p = 6.36 x 10⁻⁶; temperature, z = 3.401, p = 0.000975). Estimated marginal means correspond to the percentage values reported in Table 3. Under control conditions (15 °C, 35 ppt), 69.3 ± 5.3% of M. edulis eggs developed into normal D-stage veliger larvae. Pairwise t-tests indicated that elevated temperature increased normal development (+15.9% vs. Controls; p = 0.0132), whereas reduced salinity caused significantly less to reach the normal D-stage (–28.7% vs. Controls; p = 0.00228). Exposure to elevated temperature and reduced salinity did not cause normal development to differ from the controls (p = 0.332).

Table 3. Mean % ± SD of veliger larvae developing into Normal D-Stage, Malformed D-Stage or Trochophore Stage in different environmental treatments [Control (15 °C and 35 ppt), Elevated Temperature (18 °C and 35 ppt), Reduced Salinity (15 °C and 25 ppt) and Combined Stressor (18 °C and 25 ppt)] in the absence of pollution.

Mean % calculated from 5 replicate wells.

Neither reduced salinity or elevated temperature alone affected the proportion of malformed larvae (p = 0.116 and p = 0.723, respectively), however their combination increased malformations (+14.2% vs. Controls; p = 0.0255). Elevated temperature reduced the proportion of trochophores (–17.2% vs. Controls; p = 0.0419), while reduced salinity delayed development, with 50% of larvae remaining as trochophores (+20.8% vs. Controls; p = 0.0285). No combined effect of reduced salinity and elevated temperature was detected on the proportion of trochophores (p = 0.194).

3.2 Pollutant mixture and increased temperature

The pollutant mixture induced a concentration-dependent decline in normal development, with strong decreases observed at the Moderate and High pollutant levels (–54.1% and near-zero development vs. Controls, respectively). The Extreme pollutant mixture level caused immediate arrested development of M. edulis embryos at all environmental conditions, meaning no specimens reached any of the three tested endpoints (normal D-stage, malformed D-stage or trochophore), therefore was excluded from further analysis.

Elevated temperature mitigated pollutant toxicity (ANOVA, F_{5, 46} = 3.3, p = 0.0131; Figure 1A), and pairwise t-tests indicated increasing normal development at Trace (+19.9%, p = 0.00911), Low (+31.8%, p = 0.0057) and Environmental (+24.5%, p = 0.00874) pollutant mixture levels when compared to the same level at 15 °C. No temperature effect occurred at Moderate or High pollutant levels, where normal development was low in both.

Exploratory ANOVA indicated an interaction between temperature and pollutant mixture level on the number of malformed D-stage larvae (F_{5, 46} = 5.4, p = 0.000519, Figure 1B), driven by increased malformations at the Moderate pollutant level (18 °C: +21.4% vs. Moderate at 15 °C, p = 0.00232). A strong interaction was also detected for trochophores (F_{5, 46} = 7.4, p = 3.78 x 10⁻⁵; Figure 1C), with pairwise t-tests highlighting higher proportions of trochophores at 15 °C when exposed to Trace (p = 0.0049), Low (p = 0.00243), Environmental (p = 0.000771) and Moderate (p = 0.00168) concentrations, but no difference at High (p = 0.0942), where virtually all larvae remained in the trochophore stage.

3.3 Pollutant mixture and decreased salinity

Reduced salinity significantly exacerbated pollutant effects on normal development (ANOVA, F_{5, 45} = 6.4, p = 0.000144; Figure 2A). Pairwise t-tests indicated that normal D-stage development was consistently lower at 25 ppt for Trace (–37.1%, p = 0.000377), Low (–30.9%, p = 0.00272), Environmental (–30.6%, p = 0.00356) and Moderate (–15.2%, p = 0.0494) pollutant mixture levels. At the High pollutant mixture level, development was negligible at both salinities (p = 0.374).

A strong interaction between salinity and pollutant mixture level was observed for malformed larvae (ANOVA, F_{5, 45} = 11.1, p = 5.34 x 10⁻⁷; Figure 2B). Pairwise t-tests revealed higher malformation rates at 25 ppt for Trace (+16.6%, p = 0.00332), Low (+23.4%, p = 0.0165), and Environmental (+27.8%, p = 0.0025) pollutant mixture levels compared to the same pollution treatment at 35 ppt. At Moderate and High pollutant levels, larvae predominantly remained in the trochophore stage at both salinities (p = 0.657 and p = 0.122, respectively; Figure 2C).

3.4 Pollutant mixture and combined abiotic stressors

No three-way interaction between temperature, salinity and the pollutant mixture was detected for normal development. Exposure to combined climate stressors (18 °C and 25 ppt) produced a similar concentration-dependent response to the pollutant mixture on normal larval development as under control conditions (Figure 3A), though pairwise t-tests indicated normal development was lower at the Trace (–23.1%, p = 0.140) and Environmental (–20.4%, p = 0.0196) pollutant mixture levels.

Combined elevated temperature and reduced salinity markedly increased the toxicity of the pollutant mixture, with pairwise t-tests showing a higher proportion of malformed D-stage larvae compared to control conditions at Trace (+22.1%, p = 0.0269), Low (+39.4%, p = 0.000249) and Environmental (+39.0%, p = 0.00959) pollutant mixture levels (Figure 3B). Trochophore proportions were broadly similar between conditions however pairwise t-tests revealed lower proportions at the Low (–22.3 %, p = 0.0216) and Environmental (–18.6%, p = 0.0338) pollutant mixture levels under the combined condition (Figure 3C).

4 Discussion

The single-stressor responses observed here align closely with established physiological understanding of the effects of environmental factors on early development in bivalves. Salinity is a fundamental variable for mussel physiology across all life stages (Pourmozaffar et al., 2020) and directly regulates embryonic and larval development (Santos et al., 2020). Optimal Mytilus larval development typically occurs at 30–35 ppt with reduced salinity significantly impairing normal D-stage development and increasing abnormalities (Bayne, 1976; Hrs-Brenko et al., 1977), as corroborated by the present study with exposure to 25 ppt salinity.

Mechanistically, osmotic stress disrupts metabolism and homeostasis, elevating energy costs and reducing the capacity for growth (Barrett et al., 2022). This was reflected in the substantial proportion of larvae arrested in the trochophore stage at reduced salinity, a developmental delay associated with increased mortality risk due to prolonged exposure to environmental stressors and energetic depletion (Leal et al., 2018). Given projections of freshening in northern European estuaries (Dye et al., 2020), these findings highlight salinity decline as a primary climate–driven threat to M. edulis early development.

Temperature changes also shaped larval development, though its effects were more nuanced. Moderate warming (18 °C) enhanced normal D-stage development in non-polluted conditions, consistent with reports that elevated temperature accelerates metabolic and enzymatic processes, thereby increasing growth rates in early life stages (Sánchez Lazo and Martínez-Pita, 2012; Virgin et al., 2019; Pechenik et al., 1990). Faster progression through sensitive early life stages likely reduces exposure to stressors in the water column and depletion of energetic reserves needed for metamorphosis (Leal et al., 2018).

However, the benefits of warming are constrained by species-specific thermal limits. M. edulis is adapted to boreal and temperate waters, with a broad distribution driven primarily by water temperature but a narrower reproductive window (Bayne, 1976). As climate change warms surface waters globally (IPCC, 2023; Biguino et al., 2023), populations may be pushed toward the upper threshold of their reproductive tolerance, potentially impacting species distribution. Thus, while moderate warming can mitigate some stressors and speed up development, further increases in temperature may undermine any mitigating effects as has been observed in other Mytilus species (Sánchez Lazo and Martínez-Pita, 2012; Virgin et al., 2019).

As anticipated, increasing concentrations of the pollutant mixture reduced normal D-stage development across all environmental conditions tested. The mixture reflected typical industrialized contamination for a UK estuary, with heavy metals in the mg-range and organic pollutants in the ng-range. Iron, manganese and aluminum dominated the mixture (62.5%, 20.8% and 11.6% total metal concentration, respectively), consistent with metal loading patterns in industrialized estuaries (SEPA, 2025). Although these metals are not usually considered primary toxicants for bivalve larvae, they can modify toxicity within mixtures. At elevated exposures, essential metals such as iron and manganese can enhance reactive oxygen species production, reducing antioxidant capacity and increasing sensitivity to co-occurring pollutants (Regoli and Giuliani, 2014; Zhou et al., 2024). In addition, the precipitation of iron and manganese oxides during estuarine mixing can scavenge trace metals, reducing their dissolved concentrations and bioavailability (Singh et al., 1984). Meanwhile, aluminum becomes more bioavailable and toxic under reduced salinity, impairing ion regulation and shell formation (Gensemer and Playle, 1999).

The sensitivity of Mytilus embryos to more potent metals such as copper, zinc, cadmium and lead (Hrs-Brenko et al., 1977; Boukadida et al., 2019; Saidov and Kosevich, 2019) is well-documented, therefore it is likely that these heavy metals would contribute to abnormalities in development, even at comparatively lower concentrations than the essential metals. Copper concentrations in the Environmental, Moderate and High pollutant concentrations approached or exceeded published EC₅₀ values for Mytilus larvae (copper EC₅₀ = 17.35 ± 0.74 μg/L; Saidov and Kosevich, 2019), while zinc exceeded known EC₅₀ thresholds at Moderate and High exposures (EC₅₀ = 175 μg/L; Martin et al., 1981), suggesting a potential driving effect in the present mixture.

Although ecotoxicological data for emerging contaminants remain far more limited than for legacy pollutants (The Royal Society, 2025), several compounds have already been shown to exert embryotoxic effects on Mytilus spp. Fabbri et al. (2014) reported a dose-dependent reduction in normal development of Mytilus galloprovincialis embryos induced by PFOA and PFOS from concentrations as low as 0.1 μg/L, which is comparable to the PFOS concentration in the High level of the present pollutant mixture. In the same study, the NSAID diclofenac produced an inverted U-shaped dose-response curve, with impaired normal development and increased malformed D-veligers from 0.01 μg/L, peaking at 0.1 μg/L and having no effect at the highest concentration tested (1,000 μg/L). The Environmental level of diclofenac in the present pollutant mixture was six times higher than the reported LOEC for M. galloprovincialis, suggesting that emerging contaminants may contribute to the embryotoxic effects observed at environmentally relevant levels.

Similarly, Franzellitti et al. (2019) found that the antiepileptic drug carbamazepine and β-blocker propanolol disrupted normal larval development at concentrations as low as 0.1 μg/L and 0.01 μg/L, respectively. In the present Environmental level, carbamazepine was present at its reported LOEC and propranolol at six times higher. Although data for the remaining mixture components are limited, existing evidence indicates that even ng-level concentrations of emerging contaminants can induce embryotoxicity in early life stages of M. edulis. It is unknown and beyond the scope of the present study how or if the co-occurring pollutants found in estuarine environments will interact on the early development of M. edulis, however the study highlighted how environmental variables, temperature and salinity, can potentiate or mitigate the effects of a complex chemical cocktail.

Embryos co-exposed to reduced salinity and pollution consistently experienced amplified developmental impairment, suggesting osmotic stress reduced the larvae's tolerance to the pollutant mixture. At 25 ppt, even Trace pollutant levels significantly increased malformations, and normal development was substantially lower than at 35 ppt across Low to Moderate pollutant levels. This pattern reflects known mechanisms of salinity-pollution interactions, as reductions in salinity are known to increase metal ion uptake and toxicity (McLusky et al., 1986; Schiedek et al., 2007), reduce detoxification capacity, and exacerbate oxidative and genotoxic stress (Boukadida et al., 2024).

These findings are particularly concerning given recent freshening of the North Atlantic and increased interannual variability in the salinity of Scottish coastal waters, which is expected to be magnified by climate driven ocean circulation changes (Dye et al., 2020). Under such conditions, environmental pollution within industrialized estuaries is likely to exert stronger biological effects than predicted from single-stressor assessments. It is likely that the resulting abnormalities in development will reduce the number of larvae surviving to adulthood, reducing the recruitment potential of wild mussel beds (Lutz and Kennish, 1992).

In contrast to the reduced salinity conditions and contradicting our hypothesis, warming was shown to partially mitigate the effects of pollution at Trace to Environmental levels, increasing normal development and reducing the proportion of larvae remaining in the trochophore stage. This suggests that elevated temperature enhances metabolic and detoxification processes (Clements et al., 2018), meaning larvae progress through early life stages faster and can better cope with pollutant exposure.

The higher temperature buffered the impact of reduced salinity on normal D-stage development in the combined condition, as there was no significant difference compared to the control in unpolluted conditions, in contrast to the reduced salinity alone, and only significant differences in normal development at the Low and Environmental pollutant levels. A similar pattern has been observed in other bivalves: embryos of the brackish–water dreissenid Mytilopsis leucophaeata showed increased tolerance to reduced salinity at higher temperatures (Verween et al., 2007), suggesting that elevated temperature can partially compensate for osmotic stress during early development.

However, this mitigation capacity is undermined at moderate to high pollutant concentrations, where it became apparent that toxicity thresholds were exceeded. Moreover, warming did not mitigate the increase in malformed D-stage larvae caused by reduced salinity; instead, malformations increased sharply between Trace and Environmental concentrations, showing although larvae were growing faster, the pollution was still having a toxic effect. This contrasts with some studies reporting antagonistic interactions between warming, reduced salinity and metal toxicity (Saidov and Kosevich, 2019), suggesting that the broader pollutant mixture used here may push larvae beyond physiological tolerance limits.

The combined results indicate that climate-driven freshening, interacting with persistent industrial pollutants, poses a major threat to early development of M. edulis. However, these responses were observed in a single parental cross and therefore may not fully represent variation among mussel genotypes or spawning events. This limitation should be considered when interpreting the broader ecological implications of our findings. Blue mussels are broadcast spawners, releasing millions of gametes into the water column in mass spawning events where the likelihood of fertilization is low and unpredictable and larval mortality is variable, with estimates of around 99% (Lutz and Kennish, 1992). This means that even modest reductions in normal development can create severe recruitment bottlenecks (Lindsay et al., 2010). In industrialized estuaries, where salinity variability is increasing and pollutant mixtures remain a prevalent issue, these stressors are likely to interact which may reduce larval survival and settlement success. Future changes therefore possess considerable ecological risk, as mussel beds are ecosystem engineers and their loss has cascading consequences for water filtration, habitat structure and estuarine resilience (Schotanus et al., 2020).

5 Conclusions

While our 48-hour laboratory exposures demonstrate strong sensitivity of M. edulis embryos to pollutant mixtures under climate-related stressors, these results represent only one component of the broader suite of pressures acting on wild populations. Disruptions to normal D-stage development may contribute to recruitment limitation, but population-level declines being observed across northern Europe likely reflect multiple interacting drivers, including habitat degradation, disease, predation and longer-term climatic trends. Our findings therefore highlight a mechanistic pathway through which pollution and climate stressors could influence early life-stage survival, while acknowledging that the link to population-level change remains indirect. These interpretations are further constrained by the use of a single parental combination, selected to minimize variability associated with parental effects. As responses may differ among mussel genotypes and spawning events, further research should incorporate a wider suite of parental crosses to strengthen population-level inference.

This study demonstrates that salinity is a dominant abiotic driver of early developmental success in M. edulis, and that low salinity amplifies the toxicity of environmental pollution, with warming providing only partial, context-dependent mitigation. These results emphasize the importance of multiple stressor research for ecological risk assessment, and the chemical-mixture approach used here illustrates the value of incorporating realistic contaminant combinations to elucidate a bigger picture understanding of how current levels of pollution interact with climate change. Future work should incorporate additional stressors such as ocean acidification, turbidity and hypoxia to better reflect estuarine complexity and improve predictions of population–level responses under future climate scenarios.

Generative AI statement

The author(s) declared that Generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/focsu.2026.1918453/full#supplementary-material

Supplementary material Supplementary material at the source

Footnotes

Edited by: Kannaiyan Neelavannan, King Fahd University of Petroleum and Minerals, Saudi Arabia

Reviewed by: Andreas H. Dobler, Technical University of Munich, Germany

Sagar Gorakh Satkar, College of fishery Science, Udgir, India