Unraveling the Relationship between the Concentrations of Hydrophobic Organic Contaminants in Freshwater Fish of Different Trophic Levels and Water Using Passive Sampling

Calibration of the Chemcatcher® passive sampler and derivation of generic sampling rates for a broad application in monitoring of surface waters

We determined sampling rates for 34 pesticides, five pesticide transformation products, and 34 pharmaceutical compounds with the Chemcatcher (CC) passive sampler in a laboratory-based continuous-flow system at 40 cm/s and ambient temperature. Three different sampling phases were used: styrene divinylbenzene disks (SDB-XC), styrene divinylbenzene reversed phase sulfonate disks (SDB-RPS), and hydrophilic lipophilic balance disks (HLB), in all cases covered with a diffusion-limiting polyethersulfone membrane. The measured sampling rates range from 0.007 L/d to 0.193 L/d for CC with SDB-XC (CC-XC), from 0.055 L/d to 0.796 L/d for CC with SDB-RPS (CC-RPS), and from 0.018 L/d to 0.073 L/d for CC equipped with HLB (CC-HLB). Comparison with sampling rates from literature enabled to derive generic sampling rates that can be used for compounds with unknown uptake kinetics such as transformations products and new compounds of interest. Field trial results demonstrate that the presently derived generic sampling rates are suitable for estimating time-weighted average concentrations within reasonable uncertainty limits. In this way, Chemcatcher passive sampling can be applied approximately to a broad range of solutes without the need for deriving compound-specific sampling rates, which enable compliance checks against environmental quality standards and further risk assessment.

Chemometers: an integrative tool for chemical assessment in multimedia environments

We propose novel chemometers - passive equilibrium samplers of, e.g., silicone - as an integrative tool for the assessment of hydrophobic organic compounds in multimedia environments. The traditional way of assessing levels of organic pollutants across different environmental compartments is to compare the chemical concentration normalized to the major sorptive phase in two or more media. These sorptive phases for hydrophobic organic compounds differ between compartments, e.g., lipids in biota and organic carbon in sediments. Hence, comparability across media can suffer due to differences in sorptive capacities, but also extraction protocols and bioavailability. Chemometers overcome these drawbacks; they are a common, universal and well-defined polymer reference phase for sampling of a large range of nonpolar organic pollutants in different matrices like biota, sediment and water. When bringing the chemometer into direct contact with the sample, the chemicals partition between the sample and the polymer until thermodynamic equilibrium partitioning is established. At equilibrium, the chemical concentrations in the chemometers can be determined and directly compared between media, e.g., between organisms of different trophic levels or inhabiting different areas, between organs within an organism or between biotic and abiotic compartments, amongst others. Chemometers hence allow expressing the data on a common basis, as the equilibrium partitioning concentrations in the polymer, circumventing normalizations. The approach is based on chemical activity rather than total concentrations, and as such, gives a measure of the "effective concentration" of a compound or a mixture. Furthermore, chemical activity is the main driver for partitioning, biouptake and toxicity. As an additional benefit, the extracts of the chemometers only require limited cleanup efforts, avoiding introduction of a bias between chemicals of different persistence, and can be submitted to both chemical analysis and/or bioanalytical profiling.

Passive water sampling and air-water diffusive exchange of long-range transported semi-volatile organic pollutants in high-mountain lakes

The concentrations of legacy and currently emitted organic pollutants were determined in the freely dissolved phase of water from six high-mountain lakes in the Pyrenees (1619-2453 m) by passive water sampling. Low-density polyethylene (LDPE) and silicone rubber (SR) sheets were exposed for three consecutive periods lasting each one year between 2017 and 2020 for the study of polychlorinated biphenyls (PCBs), organophosphate esters (OPEs), polycyclic aromatic hydrocarbons (PAHs), and other organochlorine compounds (e.g., hexachlorobenzene, HCB). HCB concentrations (1.0-14 pg L^-1) remained essentially the same as those measured with pumping systems over two decades ago in the same area. ƩPAHs (35-920 pg L^-1) were around half of those observed in the past, which agrees with reductions in European atmospheric emissions. ƩPCB concentrations (1.2-2.2 pg L^-1) were substantially lower, although unexpectedly large differences could be due to comparing yearly averages from the present study to seasonally variable (i.e., affected by snowmelt, stratification, and colloidal organic matter) episodic pumping measurements from previous studies. ƩOPEs (139-2849 pg L^-1) were measured for the first time in this area and were found at high concentrations in some sites. Concentrations of most compounds obtained with LDPE and SR samplers agreed with each other by ratios generally lower than three or four times, except for a few PAHs and OPEs. Diffusive exchange flux calculations between the atmospheric gas phase and the freely dissolved water phase revealed net deposition of pollutants from air to water, except for some OPEs and PCBs presenting equilibrium conditions, and HCB with volatilization fluxes. Atmospheric degradation fluxes of PAHs and OPEs pointed at competing removal mechanisms that support the air-to-water direction of their diffusive exchange, while PCBs and organochlorines were not affected by photodegradation. In their current state, these remote lakes accumulate many emerging and legacy pollutants subject to long-range atmospheric transport.

Accessibility investigation of semi-volatile organic compounds in indoor dust estimated by multi-ratio equilibrium passive sampling

Many semi-volatile organic compounds (SVOCs) accumulate in indoor dust, which serves as a repository for those compounds. The presence of SVOCs in indoor environments is of concern because many of them are suspected to have toxic effects. Total SVOC concentrations in the dust are generally used for exposure assessment to indoor contaminants, assuming that 100% of the SVOCs is accessible for human uptake. However, such an assumption may potentially lead to an overestimated risk related to dust exposure. We applied a multi-ratio equilibrium passive sampling (MR-EPS) for estimation of SVOC accessibility in indoor settled dust using silicone passive samplers and three particle size dust fractions, <0.25 mm, 0.25-0.5 mm, and 1-2 mm in dry and wet conditions. Equilibrations were performed at various sampler-dust mass ratios to achieve different degrees of SVOC depletion, allowing the construction of a desorption isotherm. The desorption isotherms provided accessible fractions (F_AS), equivalent air concentrations (C_AIR), dust-air partition coefficients (K_DUST-AIR) and organic carbon-air partition coefficients (K_OC-AIR). The highest F_AS were observed in the <0.25 mm dust fraction in wet conditions which is relevant for exposure assessment via oral ingestion. The highest C_AIR were estimated for several organophosphorus flame retardants (OPFRs), polycyclic aromatic hydrocarbons (PAHs) and synthetic musks. The logK_OC-AIR did not differ between dust particle sizes in dry and wet conditions but within compound groups, different relationships with hydrophobicity were observed. Equivalent lipid-based concentrations (C_L⇌DUST) calculated using available lipid-silicone partition coefficients (K_LIP-SIL) were compared with lipid-based concentrations (C_L) measured in human-related samples collected from Europeans. For hexachlorobenzene (HCB), C_L⇌DUST, and C_L were similar, indicating equilibrium attainment between environment and human samples. Lipid-based concentrations for persistent legacy contaminants were also similar but lower for PBDEs in human samples. Overall, accessibility estimation using MR-EPS in dust further contributes to human risk assessment.

Dermal Uptake is an Important Pathway for the Bioconcentration of Hydrophobic Organic Compounds by Zebrafish (Danio rerio)

For bioconcentration of hydrophobic organic compounds (HOCs), most of studies assumed that fish absorb HOCs mainly through gills but often ignored the dermal uptake. In this study, deuterated polycyclic aromatic hydrocarbons (PAHs-d₁₀, phenanthrene-d₁₀, and pyrene-d₁₀) and polychlorinated biphenyls (PCB-153) were selected to study whether zebrafish can absorb freely dissolved and dissolved organic matter (DOM)-associated HOCs through dermal uptake. The results showed that the freely dissolved PAHs and PCBs could directly enter the body of zebrafish through its skin. However, PAHs and PCB-153 associated with DOM (~ 10 kDa) could not enter zebrafish through the skin. When gill and dermal exposure coexisted, dermal uptake contributed 2.9 ~ 7.6% and 31.9 ~ 38.4% of PAHs and PCB-153 bioconcentration after exposure for 6 h, respectively. The present study demonstrates that dermal uptake is an important pathway for the bioconcentration of HOCs by fish, which should be considered when studying the toxicodynamics and toxicokinetics of HOCs in organisms.

Exploring the partitioning of hydrophobic organic compounds between water, suspended particulate matter and diverse fish species in a German river ecosystem

BACKGROUND

Bioaccumulation of hydrophobic organic compounds (HOCs) along freshwater food chains is a major environmental concern as top predators in food webs are relevant for human consumption. To characterize and manage the associated risks, considerable numbers of organisms are sampled regularly for monitoring purposes. However, ethical and financial issues call for an alternative, more generic and more robust approach for assessing the internal exposure of fish that circumvents large variability in biota sampling due to interindividual differences. Passive sampling devices (PSDs) offer a fugacity-based approach for pollutant enrichment from different abiotic environmental compartments with a subsequent estimation of bioaccumulation in fish which we explored and compared to HOC concentrations in fish as determined using traditional approaches.

RESULTS

In this study, concentrations in silicone-based PSDs applied to the water phase and suspended particulate matter (SPM) of a river polluted with HOCs were used to estimate the concentration in model lipids at thermodynamic equilibrium with either environmental compartment. For comparison, muscle tissue of seven fish species (trophic level 1.8 to 2.8) was extracted using traditional exhaustive solvent extraction, and the lipid-normalized concentrations of HOCs were determined. The PSD-based data from SPM proved to be a more conservative estimator for HOCs accumulated in fish than those from water. Body length of the fish was found to be more suitable to describe increasing accumulation of HOCs than their trophic level as derived from stable isotope analysis and might offer a suitable alternative for future studies.

CONCLUSIONS

By combining fugacity-based sampling in the abiotic environment, translation into corresponding concentrations in model lipids and body length as an indicator for increasing bioaccumulation in fish, we present a suggestion for a robust approach that may be a meaningful addition to conventional monitoring methods. This approach potentially increases the efficiency of existing monitoring programs without the need to regularly sacrifice vertebrate species.

SUPPLEMENTARY INFORMATION

The online version contains supplementary material available at 10.1186/s12302-022-00644-w.

Activation of the xenobiotic metabolism and oxidative stress response by mixtures of organic pollutants extracted with in-tissue passive sampling from liver, kidney, brain and blubber of marine mammals

We used in-tissue passive equilibrium sampling using the silicone polydimethylsiloxane (PDMS) to transfer chemical mixtures present in organs from marine mammals with lipid contents between 2.3 and 99%into in vitro bioassays. Tissues from five harbor porpoises (Phocoena phocoena), one harbor seal (Phoca vitulina) and one orca (Orcinus orca) from the North and Baltic Seas were sampled until thermodynamic equilibrium was reached. Mixture effects were quantified with cellular reporter gene bioassays targeting the activation of the aryl hydrocarbon receptor (AhR-CALUX), the peroxisome proliferator-activated receptor gamma (PPARγ-bla) and the oxidative stress response (AREc32), with parallel cytotoxicity measurements in all assays. After removing co-extracted lipids and other matrix residues with a non-destructive cleanup method (freeze-out of acetonitrile extract followed by a primary secondary amine sorbent extraction), the activation of the PPARγ and AREc32 were reduced by factors of on average 4.3 ± 0.15 (n = 22) and 2.5 ± 0.23 (n = 18), respectively, whereas the activation of the AhR remained largely unaltered: 1.1 ± 0.075 (n = 6). The liver extracts showed the highest activation, followed by the corresponding kidney and brain extracts, and the blubber extracts of the animals were the least active ones. The activation of the PPARγ by the liver extracts was reduced after cleanup by a factor of 11 ± 0.26 (n = 7) and the AREc32 activity by a factor of 1.9 ± 0.32 (n = 4). The blubber extracts did not activate the AhR up to concentrations where cytotoxicity occurred or up to an acceptable lipid volume fraction of 0.27% as derived from earlier work, whereas all liver extracts that had undergone cleanup activated the AhR. The developed in-tissue passive sampling approach allows a direct comparison of the bioassay responses between different tissues without further normalization and serves as a quantitative method suitable for biomonitoring of environmental biota samples.

Passive Sampling Helps the Appraisal of Contaminant Bioaccumulation in Norwegian Fish Used for Regulatory Chemical Monitoring

Hexachlorobenzene (HCB), listed on the Stockholm Convention on persistent organic pollutants and regulated as a hazardous priority pollutant by the Water Framework Directive (WFD), is ubiquitously distributed in the environment and assumed to mildly biomagnify in aquatic foodwebs. The proposal to include trophic magnification factors (TMFs) in the procedure for comparing contaminant levels in biota at different trophic levels (TLs) with WFD environmental quality standards requires adequate selection of TMFs. In the first step of our study, we compared two independently obtained datasets of pentachlorobenzene (PeCB) and HCB concentration ratios from passive sampling (PS) in water and in fish through routine monitoring programs in Norway to evaluate possible biomagnification. In this procedure, PeCB is used for benchmarking the bioconcentration in fish, and the observed HCB/PeCB ratios in fish are compared with ratios expected in the case of (i) HCB bioconcentration or (ii) biomagnification using published TMF values. Results demonstrate that it is not possible to confirm that HCB biomagnifies in fish species that would be used for WFD monitoring in Norway and challenges the proposed monitoring procedures for such compounds in Norwegian or European waters. In the second step, fish-water chemical activity ratios for HCB and PeCB as well as for polychlorinated biphenyls where biota and PS were conducted alongside were calculated and found to rarely exceed unity for cod (Gadus morhua), a fish species with a TL of approximately 4.

Comparing total and accessible concentrations of hydrophobic organic contaminants in sediments and suspended particulate matter in the Danube River

Contamination of aquatic ecosystems by hydrophobic organic contaminants (HOCs) is often assessed based on their concentrations in riverbed sediment and suspended particulate matter (SPM). However, total HOC concentration (C_TOT) in sediment or SPM is of limited value for evaluating the exposure of benthic or pelagic organisms. The accessible HOC concentration (C_AS) presents a useful parameter quantifying the overall pool of HOC in sediment or SPM available for fast partitioning to the water phase or biota. We applied a novel approach of ex situ sequential equilibrium partitioning with silicone elastomer sampler at a high sampler/SPM phase ratio to measure C_AS of HOC in SPM from the Danube River. We compared C_TOT and C_AS in SPM and surface layer sediment collected at the same sites to evaluate whether HOC monitoring in the two matrices provides equivalent information on environmental quality. At most sites, there was a good agreement and correlation of organic carbon (OC)-normalised C_TOT in SPM and sediment for polychlorinated biphenyls (PCBs) and the majority of organochlorine pesticides (OCPs). In contrast, C_TOT of polycyclic aromatic hydrocarbons (PAHs) in SPM were up to a factor 10 lower in SPM than in sediment. Site-specific differences of OC-normalised C_AS concentrations in SPM and sediments were observed for PCBs and OCPs, with accessibility mostly lower in SPM than in sediment. The highest accessibility in SPM was observed for PCBs, ranging between 15 and 30%. The accessibility of OCPs varied from 0 to 23%. SPM and riverbed sediment samples provide complementary but not mutually interchangeable information on HOC contamination.

pH-Dependent Partitioning of Ionizable Organic Chemicals between the Silicone Polymer Polydimethylsiloxane (PDMS) and Water

The silicone polymer polydimethysiloxane (PDMS) is a popular passive sampler for in situ and ex situ sampling of hydrophobic organic chemicals. Despite its limited sorptive capacity for polar and ionizable organic chemicals (IOC), IOCs have been found in PDMS when extracting sediment and suspended particulate matter. The pH-dependent partitioning of 190 organics and IOCs covering a range of octanol-water partition constants log K _ow from -0.3 to 7.7 was evaluated with a 10-day shaking method using mixtures composed of all chemicals at varying ratios of mass of PDMS to volume of water. This method reproduced the PDMS-water partition constant K _PDMS/w of neutral chemicals from the literature and extended the dataset by 93 neutral chemicals. The existing quantitative structure-activity relationship between the log K _ow and K _PDMS/w could be extended with the measured K _PDMS/w linearly to a log K _ow of -0.3. Fully charged organics were not taken up into PDMS. Thirty-eight monoprotic organic acids and 42 bases showed negligible uptake of the charged species, and the pH dependence of the apparent D _PDMS/w(pH) could be explained by the fraction of neutral species multiplied by the K _PDMS/w of the neutral species of these IOCs. Seventeen multiprotic chemicals with up to three acidity constants pK _a also showed a pH dependence of D _PDMS/w(pH) with the tendency that the neutral and zwitterionic forms showed the highest D _PDMS/w(pH). D _PDMS/w(pH) of charged species of more hydrophobic multiprotic chemicals such as tetrabromobisphenol A and telmisartan was smaller but not negligible. Since these chemicals show high bioactivity, their contribution to mixture effects has to be considered when testing passive sampling extracts with in vitro bioassays. This work has further implications for understanding the role of microplastic as a vector for organic micropollutants.

Equilibrium sampling of suspended particulate matter as a universal proxy for fish and mussel monitoring

Bioaccumulation of persistent and hydrophobic organic compounds in the aquatic environment puts secondary consumers, such as fish, at risk. To assess their exposure, monitoring programs with high numbers of individuals have been conducted worldwide over decades that require major efforts and raise ethical issues. This study aimed at testing suspended particulate matter (SPM) as an alternative and accessible abiotic matrix to estimate the internal exposure concentrations of such chemicals in fish and mussels. Muscle tissues of bream (Abramis brama), tissues of zebra mussels (Dreissena polymorpha) and SPM were collected from four major German rivers, Elbe, Danube, Saar and Saale, in 2018 within the national monitoring program of the German Environmental Specimen Bank. We used (i) total solvent extraction for biota samples to quantify the lipid-normalized concentrations of polychlorinated biphenyls, polycyclic aromatic hydrocarbons and hexachlorobenzene and (ii) passive equilibrium sampling of SPM to derive equilibrium partitioning concentrations in lipids and (iii) set these independent data sets into context. Since the ratio of lipid-normalized concentration / equilibrium partitioning concentration in lipids was in most cases < 1.0, SPM may serve as a conservative proxy for the internal concentration of bream and mussels, although bream of high age (i.e., older than 10 years) showed a tendency for this ratio to exceed 1.0. This observation indicates that age-dependent biomagnification can exceed the predictions based on thermodynamic equilibrium relative to SPM.

Passive sampling of chlorinated paraffins by silicone: Focus on diffusion and silicone-water partition coefficients

Short-chain chlorinated paraffins (SCCPs) are under regulation through the European Water Framework Directive and were recently classified as POPs. Consecutively, the increasing use of middle-chain chlorinated paraffins (MCCPs) becomes of growing concern. Knowledge on the occurrence of chlorinated paraffins (CPs) is still scarce particularly in water phase. To achieve sufficient method sensitivity, the passive sampling approach, acting as a relevant alternative to usual grab sampling, has been considered only very recently for the monitoring of CPs in water. The present work aimed at determining the diffusion coefficients in silicone (D_s) and the silicone-water partition coefficients (K_sw) of various CP groups, having different chlorine contents and carbon chain lengths, in four commercial CP mixtures. Log D_s (-10.78 to -10.21) was found to vary little and to be high for the groups of CPs studied. Thus, their uptake in silicone is controlled by the water boundary layer, which allows to consider the release of performance and reference compounds for in-field estimation of the sampling rate. Moreover, CPs partitioned strongly towards silicone rubbers. Both the chlorination degree and the carbon chain length of CPs cause large uncertainties in the partitioning between silicone and water (log K_sw between 4.85 and 6.30), indicating that instead of an average value, differentiated K_sw should be used to estimate aqueous CPs more accurately. Even so, the probable influence of chlorine atoms position on polarity and partitioning may be an argument for favoring sampling in the kinetic stage.

Kinetics of Equilibrium Passive Sampling of Organic Chemicals with Polymers in Diverse Mammalian Tissues

Equilibrium passive sampling employing polydimethylsiloxane (PDMS) as a sampling phase can be used for the extraction of complex mixtures of organic chemicals from lipid-rich biota. We extended the method to lean tissues and more hydrophilic chemicals by implementing a mass-balance model for partitioning between lipids, proteins, and water in tissues and by accelerating uptake kinetics with a custom-built stirrer that effectively decreased time to equilibrium to less than 8 days even for a homogenized liver tissue with an only 4% lipid content. The partition constants log K_lipid/PDMS between tissues and PDMS were derived from measured concentration in PDMS and the mass-balance model and were very similar for 40 neutral chemicals with octanol-water partition constants 1.4 < log K_ow < 8.7, that is, log K_lipid/PDMS of 1.26 (95% CI, 1.13-1.39) for the adipose tissue, 1.16 (1.00-1.33) for the liver, and 0.58 (0.42-0.73) for the brain. This conversion factor can be applied to interpret chemical analysis and in vitro bioassays after additionally accounting for a small fraction of coextracted lipids of <0.7% of the PDMS weight. PDMS is more widely applicable for passive sampling of mammalian tissues than previously thought, both, in terms of diversity of chemicals and the range of lipid contents of tissues and, therefore, an ideal method for human biomonitoring to be combined with in vitro bioassays.