The influence of temperature and conductivity on metabolism and Per- and Polyfluoroalkyl Substance (PFAS) bioconcentration in Bluegill (Lepomis macrochirus)

TL;DR

This study explores how temperature and water conductivity affect bluegill fish metabolism and their accumulation of harmful PFAS chemicals.

Contribution

The study identifies how temperature and conductivity influence PFAS bioconcentration and metabolic rates in bluegill.

Findings

PFOS and PFHxS accumulation in the liver was higher at 20 °C compared to 25 °C.

Contaminant elimination was faster at higher temperatures, linked to increased metabolic rates.

Resting and maximum metabolic rates increased with temperature, regardless of conductivity.

Abstract

Aquatic organisms may be exposed to multiple concurrent stressors, including anthropogenic contaminants and alterations to abiotic parameters such as temperature and conductivity. Amid global climate change, elucidating the interactions between chemical and environmental factors is a priority. In recent years, per- and polyfluoroalkyl substances (PFAS) have emerged as contaminants of global concern owing to their persistence, bioaccumulation potential, and toxicity. While limited studies have documented the influence of temperature and conductivity on PFAS bioaccumulation in fish, the underlying physiological drivers associated with changes to uptake and elimination remain poorly understood. Therefore, the present study aimed to determine the influence of temperature and conductivity on metabolic parameters and bioconcentration of an environmentally relevant PFAS mixture in bluegill, Lepomis macrochirus. Fish were exposed to a mixture of perfluorooctane sulfonate (PFOS), perfluorohexane sulfonate (PFHxS), perfluorooctanoic acid (PFOA), and perfluorohexanoic acid (PFHxA) at 20 and 25 °C and conductivity levels of 300, 600, and 1200 µS/cm for a 32-day uptake and 32-day depuration period. Additionally, we measured individual resting metabolic rate (RMR), maximum metabolic rate (MMR), and aerobic scope under the same temperature and conductivity conditions. Overall, PFOS and PFHxS accumulation in the liver were consistently higher in fish at 20 °C compared to 25 °C, which was linked to more rapid contaminant elimination at higher temperatures. Resting and maximum metabolic rates increased with temperature, independent of conductivity. These results will facilitate the development of robust bioaccumulation models to enhance predictive modeling of PFAS accumulation and aid in understanding the physiological drivers of fish contaminant accumulation.

The online version contains supplementary material available at 10.1007/s10646-026-03027-6.

Figures

Uptake and depuration of PFAS in bluegill (*Lepomis macrochirus*) exposed to environmentally relevant concentrations of (A) PFOS, (B) PFHxS, (C) PFOA, and (D) PFHxA at three different conductivity levels (300, 600, and 1200 µS/cm) and 20 °C (green) and 25 °C (red). Nominal exposure concentrations were 1, 0.600, 0.170, and 0.250 µg/L for PFOS, PFHxS, PFOA, and PFHxA, respectively. The experiment lasted for a total of 64 days with a 32-day uptake and depuration period. The transition to depuration is noted by the vertical dashed line. The dots show mean values in liver tissue on a wet weight (ww

Comparison of PFHxS elimination in bluegill, *Lepomis macrochirus*, exposed at 300 µS/cm at 20 °C (A) and 25 °C (B). Concentrations are those in the liver on a wet weight (ww) basis from day 32 (start of the depuration period) to day 64 of bioconcentration study. Each point represents a measurement of PFHxS in fish liver with triplicate measurements at each timepoint. The line represents the fitted depuration curve for PFHxS using a first-order decay model

Metabolic performances of juvenile bluegill in three temperatures and three ecologically relevant freshwater salinization conditions. (A) Resting metabolic rate (RMR). (B) Maximum metabolic rate (MMR). (C) Absolute Aerobic Scope (AAS = MMR - RMR). (D) Factorial Aerobic Scope (FAS = MMR/RMR). Each individual value is plotted as small, faded symbols; the predicted marginal means, and error are plotted in the dark large symbols. The letters show the difference between temperatures. The shape is conductivity. Body mass scaling exponents (b) are shown on the panel. All data were normalized to the m

Full text

Introduction

Fish may be exposed to a growing number of concurrent stressors, including changing temperature regimes, freshwater salinization, and anthropogenic chemicals such as per- and polyfluoroalkyl substances (PFAS) (Velasco et al. 2019). Exposure to multiple stressors with different modes of action can result in additive, synergistic, or antagonistic effects (Todgham and Stillman 2013). These changes in abiotic conditions can influence the bioaccumulation and toxicity of contaminants such as PFAS (Lewis et al. 2022); however, the influence of environmental factors and potential underlying mechanisms on contaminant bioaccumulation are rarely tested under ecologically relevant multiple stressor conditions.

PFAS comprise a diverse group of several thousand anthropogenic chemicals extensively used in industrial and consumer products since the 1950s for food packaging, waterproofing of fabrics, firefighting foams, cosmetics and others (Glüge et al. 2020). PFAS are persistent and have a high affinity to lipids and proteins resulting in their high bioaccumulation potential in fish, terrestrial wildlife, and humans (Buck et al. 2011; Burkhard 2021). As such, regulatory agencies including the United States Environmental Protection Agency (US EPA) and the European Chemical Health Agency (ECHA) are developing limits for the control of PFAS, including guidance on the consumption of PFAS-contaminated foodstuff, often fish (European Food Safety Authority, 2022; Young et al. 2022). Quantifying bioaccumulation of PFAS is of particular importance in supporting the development of regulatory thresholds, both from the context of human consumption and the potential for adverse effects of PFAS on fish.

Fish have been the focus of numerous studies of PFAS bioaccumulation due to their importance in human exposure to PFAS (Barbo et al. 2023; Pasecnaja et al. 2022; Rüdel et al. 2022), and utility in long-term PFAS monitoring programs (Valsecchi et al. 2021). Broadly used bioaccumulation factors (BAFs) and bioconcentration factors (BCFs) in fish have demonstrated significant variability across and within species and PFAS tested (Burkhard 2021). For most PFAS, phospholipid partitioning and protein-binding mechanisms are important drivers of bioaccumulation in fish, with serum protein binding known to be influenced by carbon chain length (Sun et al. 2022; Zhong et al., 2019). Additionally, tissue bioaccumulation varies substantially even among fish species collected from the same location in the field (Munoz et al. 2017; Brown et al. 2023; Lanza et al. 2017). Variations can be attributed to the different PFAS physical and chemical properties, including perfluorocarbon chain length and polar functional group as well as precursor PFAS transforming to other PFAS within the environment and biota (Burkhard 2021; Lewis et al. 2022). In addition, other ecologically relevant factors like temperature, conductivity, and exposure concentrations can affect the magnitude of bioaccumulation (Lewis et al. 2022). Furthermore, fish physiological parameters including metabolic rates, gill ventilation, and elimination rates have the potential to influence accumulation of PFAS. Identifying how the interaction between parameters such as conductivity and temperature as well as fish physiology affects PFAS bioaccumulation in fish is imperative to inform ongoing monitoring and PFAS risk assessment efforts. Based on available literature and to the best of our knowledge, these complex interactions have not yet been studied under environmentally relevant conditions. In addition, many of the existing bioconcentration studies exposed fish to PFAS in singular (Martin et al. 2003; Jeon et al. 2010; Vidal et al. 2019), which is not representative of typical environmental exposures. Aquatic organisms are typically exposed to PFAS in complex mixtures reflecting the diverse sources and transport pathways of this class of contaminants. In addition, toxicokinetic studies utilizing mixture exposures have found differences in uptake and elimination of PFAS compared with singleton exposures (Golosovskaia et al. 2024).

At present, the underlying physiological mechanisms associated with the observed differences in PFAS bioaccumulation among and within species are poorly understood. Knowledge about physiological performance and PFAS bioaccumulation in the same species under varying abiotic conditions will improve our ability to predict field-based PFAS accumulation in fish (Golosovskaia et al. 2024; Sun et al. 2022). For example, metabolic rates that are related to gill ventilation may be linked with PFAS contaminant uptake and elimination rates through the gills (Yang et al. 2000). Consistent with high interindividual variation in PFAS bioconcentration (Macorps et al. 2022), metabolic rates also vary substantially between individuals. Additionally, metabolic rates typically increase with temperature (Clark et al. 2013; Killen et al. 2010) and can change with specific conductance in freshwater fish (Zikos et al. 2014).

We aimed to investigate the links between metabolic performance and uptake and elimination of a PFAS mixture (PFOS, PFHxS, PFHxA, and PFOA) in juvenile bluegill, Lepomis macrochirus, under ecologically relevant temperatures and freshwater salinization conditions. Specifically, we measured resting metabolic rates (RMR; the metabolic rates in resting, non-digesting fish), maximum metabolic rates (MMR; exercise-induced maximum metabolic rates) and the difference between MMR and RMR or aerobic scope (AS = MMR - RMR) that represents an individual’s aerobic capacity to perform vital functions like growth, reproduction, and feeding (Clark et al. 2013). Additionally, we performed 64-day (32-day uptake and 32-day depuration) bioconcentration studies in L. macrochirus exposed to an environmentally relevant PFAS mixture at varying temperature (20 and 25 °C) and conductivity conditions (300, 600, and 1,200 µS/cm) representative of global climate change and freshwater salinization, respectively. Concentrations of PFAS were measured in bluegill liver only owing to the known elevated bioconcentration of PFAS in this tissue (Jeon et al. 2010) and to facilitate comparison with other PFAS bioconcentration studies focusing on liver bioconcentration (Jeon et al. 2010; Martin et al. 2003; Vidal et al. 2019). The PFAS mixture used was based on repeated field measurements in aqueous film forming foam (AFFF)-contaminated surface water from two US Department of Defense (DoD) sites in the eastern US and is known to be representative of PFAS mixtures occurring in contaminated sites nationwide (East et al.2021). The data generated from this study will advance mechanistic understanding behind the variation in fish PFAS bioconcentration under ecologically relevant multiple stressor conditions.

Methods

Ethical Statement

All procedures were approved by Towson University’s Institutional Animal Care and Use Committee (IACUC) protocol 2086 and EA protocol 2022.06-1. All experiments were conducted at EA Engineering, Science and Technology Inc., PBC’s Ecotoxicology Laboratory (Hunt Valley, Maryland US).

Standards and Reagents

Standards of PFOS, PFHxS, PFHxA, and PFOA in methanol were purchased from Sigma Aldrich (St Louis, Missouri, USA). The stock solutions of PFAS mixture containing PFOS, PFHxS, PFHxA, and PFOA, were prepared in deionized water, allowed to mix overnight on a shaker table, and stored at room temperature prior to use. Further details on the PFAS used for this study are provided in Table 1. Sodium chloride was purchased from Sigma Aldrich. Tricaine methanesulfonate (MS-222), a fish anesthetic, was obtained from Sigma Aldrich. A solution of buffered 1 g/L MS-222 was used to euthanize fish and a buffered 0.08 g/L MS-222 solution was used to anesthetize fish.

Table 1. Physicochemical properties of the four PFAS used in the study Full Chemical NameAcronymCASChemical FormulaMolecular Weight(g/mol)LogKowPurityPerfluorooctane sulfonatePFOS1763-23-1C_8_HF_17_O_3_S500.135.6197.9%Perfluorohexane sulfonatePFHxS355-46-4C_6_HF_13_O_3_S400.122.2098%Perfluorooctanoic acidPFOA335-67-1C_8_F_15_O_2_414.073.10100%Perfluorohexanoic acidPFHxA307-24-4C_6_HF_11_O_2_314.052.8598%LogKow value provided are experimental averages from US EPA CompTox dashboard

Fish Husbandry and Acclimations

Juvenile bluegill (bioassay grade, length range = 59–75 mm, average length = 65.6 ± 4.18 mm [mean ± standard deviation], weight range = 2.24–4.89 g, average weight = 3.32 ± 0.701 g) were obtained from Osage Catfisheries (Osage Beach, Missouri, USA) and acclimated to treatment temperature and control (300 µS/cm) conductivity conditions for 7 days prior to use in bioconcentration studies. Fish for metabolic rate measurements were kept at 12 °C prior to acclimation, which was adjusted at a rate of 2 °C/h to reach the target acclimation temperature. This rate of temperature change is significantly lower than what bluegill are typically exposed to in the environment (Winter et al. 2018). Changes in conductivity levels were acute. The experiments were staggered by temperature (one temperature at a time, all three conductivity levels concurrently) and conducted between June 2023 and January 2024. Partial water changes and water quality were conducted at least once weekly throughout the experiments: i) weekly pre- and post-renewal for the bioconcentration study (Orion Star ™ A329 Portable Multiparameter Meter for pH, temperature, DO and Conductivity; Orion ™ Versa Star Pro ™ Benchtop pH Meter for ammonia), and 2) weekly tests (API 5-in-1 test strips and API freshwater master test kit; Chalfont, PA, USA). The temperatures were kept within ± 1 °C of target temperature and ± 100 µS/cm target conductivity throughout all acclimations (summarized in Supplemental Table S1) with a photoperiod of 12 h L:12 h D. Throughout the study, all fish were fed daily with commercially available feed Aquatic Nutrition Giant Fish Grower (Eustis, Florida, USA) at a minimum of 1.5% body weight (considered ad libitum, leftover food was removed from the tanks between feedings). Feed was screened for background PFAS contamination prior to beginning the experiment. For feed, none of the quantitated PFAS were detected above the limit of quantification, with trace concentrations of PFOA and PFDA (< 1 ng/g) detected in 2 out of 3 feed replicates analyzed.

Bioconcentration Study

Bioconcentration studies were designed following the recommendations of ASTM E1022 (American Society for Testing and Materials [ASTM] 2022) and previous studies of PFAS bioaccumulation in fish (Jeon et al. 2010). Fish were exposed to PFAS mixtures for a 32-day uptake phase, followed by a 32-day depuration period in dechlorinated tap water. Bioconcentration studies were conducted at three conductivity levels (300, 600, and 1200 µS/cm) and two temperatures (20 and 25 °C). The selected conductivity treatments (300, 600, and 1200 µS/cm) are representative of freshwater salinization occurring in the northeastern US over the past 30 years (Utz et al. 2022) and predicted future conditions (Olson, 2019). Temperature conditions represent typical spring spawning conditions for bluegill (~ 20 °C, Yamamoto & Shiah, 2013), and projected stream temperature increases over the next century (Caldwell et al. 2015). Fish were acclimated to temperatures for 7 days prior to initiating the experiment. The target PFAS exposure concentrations were based on levels measured in PFAS-impacted Department of Defense (DoD) sites conducted as part of the field component of a larger project (Brown et al. 2023). Nominal PFAS concentrations were as follows: 1000, 600, 250, and 170 ng/L for PFOS, PFHxS, PFHxA, and PFOA, respectively. Additionally, a control exposure group housed in PFAS-free dechlorinated tap water at 300 µS/cm was included at each temperature.

Fish at each temperature and conductivity treatment were housed in a semi flow-through exposure system with a reservoir and three replicate tanks (n = 40 fish per tank, n = 120 fish per treatment, approximate loading rate of 0.8 g/L (American Society for Testing and Materials [ASTM], 2022;150 L polypropylene tank; 38 in L x 22 in W x 15 in H filled to 100 L, total volume per treatment 350 L). Experiments were conducted in a temperature-controlled room set at the target temperature (20 or 25 °C). Water was pumped from the reservoir to replicate tanks at a continuous rate of 2 volume additions per day, the flow rate was checked and adjusted weekly. The flow rate ranged from 800 to 1200 mL min during the study. PFAS-containing solutions (or control water) were added to the header and circulated through each tank, with each tank equipped with an outflow valve to return water to the header. The exposure system did not include any filtration, with feces and excess food removed by hand. Previous studies have indicated that PFAS readily bind to dissolved organic matter in bioconcentration studies, with increasing organic material known to reduce both measured uptake and elimination constants (Wen et al. 2016). All tanks were aerated throughout the exposure and covered with a plastic lid. During the 32-day uptake phase, approximately 80% of the header tank capacity (equivalent to 200 L) was drained and renewed on days 5, 12, 19, 26, and 32. During the depuration period, water was renewed using the same procedure and dechlorinated water only on days 36, 43, 50, and 57. The day prior to water renewals, four 200 L solutions were made with dechlorinated tap water, adjusted to each conductivity with sodium chloride, and PFAS mixture stock was added. The same PFAS mixture stock was used for the entire study with confirmed PFAS concentrations. Renewal solutions were allowed to aerate and adjust to treatment temperature overnight. During renewals, approximately 200 L (~ 80%) of water was removed from each semi flow-through exposure system and replaced with fresh solution. After the 32-day PFAS uptake phase, fish were removed from tanks and placed in holding tanks with PFAS free water (at the target treatment conductivity and temperature). A new flow-through system filled with dechlorinated tap water adjusted with sodium chloride for each conductivity was installed during the holding period. Fish were then placed back in their housing tanks for the 32-day PFAS depuration phase. The same water renewal protocol (except for PFAS addition) was followed until the end of the trial.

Throughout the experiment, fish were sampled on days 0, 2, 4, 8, 16, 24, and 32 during the uptake period, and days 2, 4, 8, 16, 24 and 32 during the depuration period (equivalent to days 34, 36, 40, 48, 56, and 64 of the entire experiment). At each time point, three or four fish were randomly removed from each tank (totaling 9 or 12 fish per treatment), weighed and measured, euthanized in MS-222, and livers were dissected for subsequent PFAS analysis. Fish wet weight was measured using an OHAUS Adventurer balance (OHAUS Corporation, Parsippany, New Jersey, US) and total length was assessed to provide a measure of growth at each time point. Survival was assessed daily. To meet sample mass requirements (200 mg), liver samples were pooled from 3 to 4 fish randomly for a total of 3 tissue samples per treatment per timepoint. Water samples (1.5 mL) to determine PFAS concentrations were taken from the reservoir after weekly renewal on days 0, 5, 12, 19, 26, and 32 during the uptake period, and days 8, 15, 22, 29, and 32 during the depuration period (equivalent to days 40, 47, 54, 61, and 64 of the entire experiment). All water and liver samples were stored at -20 °C, respectively, prior to shipment on ice to Purdue University for subsequent chemical analysis.

Metabolic Rate Study

This part of the study examined metabolic rates in individual fish acclimated to each of three temperatures (16, 20, 25 °C) in combination with each conductivity condition (300, 600, 1200 µS/cm). An additional temperature treatment (16 °C) was included for the metabolic rate study but not assessed for bioconcentration due to the presence of disease in the batch of fish designated for the bioconcentration study at this temperature. Fish for metabolic rate measurements were kept in 20 L tanks (2 tanks per treatment, n = 6 fish/tank). The acclimations were staggered, with one temperature treatment conducted at a time. Each tank was supplied with an aquarium filter (Aqueon, Central Aquatics, Franklin WI, USA) to maintain water quality, leftover food and feces were removed by hand. Tanks from the 25 °C experiment were equipped with a submersible 50 W heater to maintain target temperatures. The water change protocol followed the same procedure as described above, except there was no PFAS addition to any tanks and renewal water was made in 5-gallon jugs. Fish were acclimated to their assigned treatment for at least 14 days before testing. Feeding was withheld the day before the respirometry trial.

Oxygen consumption rates, a proxy for metabolic rates (MO_2_), were measured using intermittent flow respirometry following published guidelines (Killen et al. 2021). The respirometry system contained 4 respirometry chambers (484 mL; Loligo Systems, Viborg, Denmark), each equipped with two flow-controlled pumps (Eheim universal; EHEIM GmbH & Co. KG. Deizisau, Germany), and an oxygen sensor (flow through oxygen cell connected to fiber optic cable; Loligo Systems, Viborg, Denmark). One pump recirculated water past the oxygen sensor and one pump was intermittently turned on and off to flush the respirometer and replenish oxygen levels. The temperature was maintained within ± 1 °C of target temperature and monitored using Pt100 temperature sensor (Loligo Systems, Viborg, Denmark). Fiber optic cables and temperature sensor were connected to Witrox 4 oxygen meter and data were recorded using WitroxView 2 software (Loligo Systems, Viborg, Denmark).

To elicit maximum metabolic rate, individual fish were manually chased by hand for 3-minutes which was followed by 1 min air exposure. Immediately after air exposure, fish were placed in the respirometry chamber, and their metabolic rates were recorded (MMR). Fish recovered in respirometers overnight on an automated 20-minute measurement cycle yielding > 60 MO_2_ measurements (flush: measurement cycles were changed with temperature treatment; oxygen was maintained > 80% air saturation). After the respirometry trial, each fish was weighed, implanted with a visible fluorescent Elastomer tag (Northwest Marine Technology, Inc, Anacortes, WA, USA) and returned to their housing tanks. All chases were performed between 9:00 and 13:00 h. Background respiration was measured in empty respirometry chambers before and after each trial. The water in the respirometry tub was changed fully before each trial and conductivity (µS/cm) was measured before and after the trial.

Chemical Analyses

All native and mass-labeled PFAS were purchased from Wellington Laboratories and all solvents used in sample preparation and measurement were liquid chromatography mass spectrometry (LC/MS) grade. Samples were extracted following methods reported by Choi et al. (2023) and quality control procedures followed U.S. EPA Method 1633 (United States Environmental Protection Agency [USEPA], 2024). Detailed information of extraction and LC-MS measurement conditions are described in Section S2 and Table S2 of the Supplementary Information (SI).

Data and Statistical Analyses

In the bioconcentration study, PFOS detections in fish not exposed to PFAS mixtures were addressed by subtracting control PFOS concentrations from the PFOS concentrations measured in the treated fish residues. Detections of other target PFAS were sporadic and at low concentrations (PFHxS range in control fish: < LoD – 0.529 ng/g ww and < LoD – 1.41 ng/g ww at 20 and 25 °C respectively; PFOA range in control fish: < LoD – 0.480 ng/g ww and < LoD – 0.0104 ng/g ww at 20 and 25 °C respectively; PFHxA range in control fish: < LoD – 0.211 ng/g ww and < LoD – 4.05 ng/g ww at 20 and 25 °C respectively); thus, no correction was performed. Calculated toxicokinetic parameters for each PFAS and temperature/conductivity treatment are presented and discussed in the SI Section S3. Due to significant variability in PFAS concentrations within uptake and depuration periods for select PFAS and treatments, analyses and discussion are focused on overall trends across temperature and conductivity treatments rather than a detailed analysis of toxicokinetic parameters. Bioconcentration factors (BCFs) were calculated as the day-32 measured liver concentrations divided by the average aqueous concentrations over the 32-d uptake period for each compound (BCF = \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\frac{d32\:Liver\:Concentration\:(\boldsymbol{\mu\:}\boldsymbol{g}/\boldsymbol{k}\boldsymbol{g})}{Average\:Aqueous\:Concentration\:(\boldsymbol{\mu\:}\boldsymbol{g}/\boldsymbol{L})}$$\end{document} ). BCF values could not be calculated for PFHxA, which was consistently measured near the detection limit in liver tissue. BCFs were also estimated using the calculated toxicokinetic parameters given in the SI Section S3. All measured aqueous PFAS concentrations are summarized in SI Tables S3, S4, and S5. All tissue data are summarized in SI Tables S6, S7, and S8. All tissue concentrations are reported in ng/g wet weight (ww), unless otherwise specified. Fish growth at each timepoint relative to day 0 measurements was analyzed using analysis of variance (ANOVA) with treatment (salinity) and day of measurement used as factors. Assumptions of ANOVA were tested using a Shapiro-Wilk and a Bartlett’s test for normality of residuals and homogeneity of variances, respectively.

Metabolic rate estimates for each measurement cycle were obtained from linear decrease in dissolved oxygen content (mgO_2_ L^− 1^) over time (min) (fitted simple linear regressions). All linear regressions were visually assessed for quality and linearity. Only regressions with R^2^ > 0.9 were used for final analysis. Oxygen uptake (MO_2_, mgO_2_ min^− 1^) was calculated following equation: MO_2_ = [(mfishV) – (mbackgroundV)], where m is the slope of each regression, and V is the volume of the respirometer (L). MMR was the highest MO_2_ recorded over at least 120 s period in the first measurement cycle (immediately after chase and air exposure) or any time during the trial (18% fish reached MMR during the change in photoperiod). Resting metabolic rate (RMR) was calculated as the mean of the 10 lowest estimated MO_2_ values after excluding the five lowest values from the entire trial (Chabot et al. 2016). The first 60 s of each measurement were excluded to account for mixing or wait period. The absolute aerobic scope was calculated as AAS = MMR - RMR and factorial aerobic scope was calculated as FAS = MMR / RMR.

To compare MMR, RMR, AAS, and FAS between different treatments, simple linear models where constructed where metabolic rate performance was a response variable (mg O_2_ min^− 1^), conductivity treatment (µS/cm; categorical variable) and temperature treatment (ºC; categorical variable) were predictor variables in interaction, and body mass (kg; continuous variable) was a covariate. The significance of each predictor was tested using Type II ANOVA, and Tukey’s post hoc test was used to identify differences in significant predictors.

All statistics were performed using R Studio v4.3.2 (2023.10.31). The packages nls, aomisc, and nlstools were used to perform modeling, test assumptions, and identify goodness-of-fit for toxicokinetic models. Metabolic rate measurements were analyzed using custom-written library in R (https://github.com/kraskura/AnalyzeResp). The packages used for metabolic rate analyses included: lme4, car, emmeans, merTools. The visualizations were created using packages ggplot2 and cowplot.

Results

Bioconcentration Study

Water Quality and Measured Water Concentrations

Water quality measurements including temperature, pH, conductivity, and dissolved oxygen taken over the course of the bioconcentration experiment are given in SI Table S1. Overall, measured temperature and conductivity values were close to the target across all treatments, with average temperatures ranging from 19.1 to 19.3 °C (20 °C treatment) and 24.3–24.7 °C (25 °C treatment), and the average conductivity ranging between 343 and 372 µS/cm, 664 and 682 µS/cm, and 1220 and 1200 µS/cm for the 300, 600, and 1200 µS/cm treatments, respectively (Table S1). Average aqueous concentrations for each PFAS during the uptake period are summarized in Table 2 (detailed information in SI Tables S4 and S5). During depuration, measured concentrations were often below the limit of detection (LOD) or only one of the replicates had quantifiable concentrations (detailed in SI Tables S3 and S4).

Fish Mortality and Growth

Mortality was low across all treatments, with average survival of 94%, 93.3%, and 91.3% for the 300, 600, and 1200 µS/cm treatments, respectively, at 20 °C. For the 25 °C treatment, survival was 99.3%, 98%, and 98% for the 300, 600, and 1200 µS/cm treatments, respectively. Control fish not exposed to PFAS mixtures at 20 and 25 °C had comparable survival of 94% and 98% respectively, exceeding acceptability criteria for commonly used fish bioconcentration tests of > 90% survival in controls (USEPA, 2016). In terms of growth, day of experiment and temperature significantly impacted growth (ANOVA, F_13, 1079_ = 18.7, p < 0.05 and ANOVA, F_1, 1079_ = 795, p < 0.05 for day and temperature, respectively), with fish at 25 °C exhibiting faster growth compared to 20 °C. Conductivity treatment and PFAS exposure had no significant effect on bluegill growth (ANOVA, p > 0.05). Average weight data for all fish at both temperatures is given in SI Figure S1.

PFAS Accumulation and Elimination

PFAS bioaccumulation and elimination varied by compound, temperature, and conductivity (Fig. 1; Tables 3 and 4). For PFOS, concentrations in bluegill liver generally increased over the uptake period for all temperatures and conductivity levels, with maximum concentrations in fish exposed at 20 °C approximately twice of those in the 25 °C treatment (Fig. 1). Measured control-corrected 32 d liver PFOS concentrations in the 20 °C treatment were 257 ± 57.7, 187 ± 23.2, and 227 ± 11.2 (average ± standard deviation) ng/g in the 300, 600, and 1200 µS/cm treatments, respectively, compared with 83.5 ± 33.0, 57.5 ± 4.14, and 88.5 ± 19.2 ng/g in the 300, 600, and 1200 µS/cm treatments at 25 °C, respectively. Similar trends were observed for PFHxS, with higher 32 d concentrations in the liver of fish exposed at 20 °C across all conductivity levels (PFHxS range: 2.79–7.34 ng/g), compared to 25 °C (range: 1.58–2.23 ng/g). For PFOA, measured liver concentrations were low across all temperatures and conductivity levels, with maximum average concentrations of 0.742 ± 0.529 ng/g and 0.311 ± 0.118 ng/g at 20 °C and 25 °C, respectively. Across both temperatures and all conductivity levels, liver PFOA concentrations peaked within the first 10 days of exposure and generally declined throughout the remainder of the uptake period (Fig. 1), with significant variability observed among independent tissue samples at a given time point, and within temperature and conductivity treatments. Lastly, PFHxA tissue concentrations were the lowest and had the least consistent bioconcentration trend over time. Somewhat unexpected and inconsistent detection of low PFHxA concentrations between 0.5 and 3 ng/g ww were measured in fish at day 0 for the 25 °C, and in 25 °C fish at 1200 µS/cm across the uptake and depuration phases, while the concentrations at all other timepoints and in 20 °C fish were around the limit of detection (Fig.1). For the 20 °C treatment, PFHxA concentrations were consistently around the limit of detection throughout the uptake and depuration periods (Fig. 1).

Fig. 1. Uptake and depuration of PFAS in bluegill (Lepomis macrochirus) exposed to environmentally relevant concentrations of (A) PFOS, (B) PFHxS, (C) PFOA, and (D) PFHxA at three different conductivity levels (300, 600, and 1200 µS/cm) and 20 °C (green) and 25 °C (red). Nominal exposure concentrations were 1, 0.600, 0.170, and 0.250 µg/L for PFOS, PFHxS, PFOA, and PFHxA, respectively. The experiment lasted for a total of 64 days with a 32-day uptake and depuration period. The transition to depuration is noted by the vertical dashed line. The dots show mean values in liver tissue on a wet weight (ww) basis of three sampled fish per timepoint, the error bar is standard error of the mean

Calculated bioconcentration factors generally declined with increasing conductivity and temperature (Tables 3 and 4), though results were compound specific. The BCFs for PFOS were the highest, ranging from 285–442 L/kg at 20 °C (Table 3), which was higher compared to 25 °C, wherein BCFs ranged from 155–182 L/kg across all conductivity levels (Table 4). Only PFOA followed a clear trend of declining BCF values with increasing conductivity at both temperatures (Tables 3 and 4). Overall, conductivity effects were less pronounced than temperature effects on PFAS uptake. At 20 °C, higher maximum concentrations of PFOS and PFHxS were observed in the 300 µS/cm treatment relative to 600 and 1200 µS/cm; however, this trend was not observed in the 25 °C treatment, suggesting possible interactive effects between temperature and conductivity (Fig. 1).

Elimination patterns differed among the individual PFAS, temperature, and conductivity conditions, with all PFAS, except for PFOS and PFHxA in the 1200 µS/cm treatment, nearly fully eliminated within the 32-day depuration period (Fig. 1). PFOS concentrations initially decreased in the liver after the transition to depuration (d32), followed by a slight increase and minimal subsequent depuration across all treatments. In contrast, PFHxS showed a clear depuration across all conditions, with the rate of elimination clearly influenced by temperature (Fig. 2). Elimination constants for PFHxS (Table S7) were notably higher in fish exposed at 25 °C (range: 99.4–179 L/d) compared to 20 °C (range: 56.1–71.5 L/d), regardless of conductivity. For PFOA, elimination appeared highly influenced by temperature, with rapid depuration to concentrations below detection limits observed in fish exposed at 25 °C, but more variable concentrations during the depuration period at 20 °C (Fig. 1). No consistent trends in PFHxA elimination were observed across either temperature or conductivity.

Fig. 2. Comparison of PFHxS elimination in bluegill, Lepomis macrochirus, exposed at 300 µS/cm at 20 °C (A) and 25 °C (B). Concentrations are those in the liver on a wet weight (ww) basis from day 32 (start of the depuration period) to day 64 of bioconcentration study. Each point represents a measurement of PFHxS in fish liver with triplicate measurements at each timepoint. The line represents the fitted depuration curve for PFHxS using a first-order decay model

Table 2. Average aqueous PFAS concentrations during the bioconcentration studyTemperature(ºC)Salinity (µS/cm)PFOS (µg/L)PFHxS(µg/L)PFOA(µg/L)PFHxA(µg/L)203000.581 ± 0.4000.483 ± 0.2770.119 ± 0.4290.229 ± 0.0766000.656 ± 0.3590.368 ± 0.0510.117 ± 0.3200.220 ± 0.03112000.580 ± 0.4010.446 ± 0.2130.116 ± 0.0130.229 ± 0.055253000.568 ± 0.3090.573 ± 0.1850.120 ± 0.0240.333 ± 0.0526000.473 ± 0.2250.585 ± 0.2950.113 ± 0.0320.310 ± 0.05412000.463 ± 0.2140.642 ± 0.4550.170 ± 0.0180.339 ± 0.043Average values are calculated across triplicate samples taken at days 0, 5, 12, 19, and 26 of the 32-day uptake period for the bioconcentration study. Values are provided ± one standard deviation

Table 3. Average PFAS concentrations in Bluegill liver during the 32-day uptake period at 20 °CDay of ExperimentConductivity (µS/cm)PFOS Concentration (ng/g ww)PFHxS Concentration (ng/g ww)PFOA Concentration(ng/g ww)PFHxA Concentration (ng/g ww)030037.0 ± 15.2BDLBDL0.0294 ± 0.026*259.1 ± 14.41.54 ± 0.1950.496 ± 0.1650.129 ± 0.150486.9 ± 54.71.68 ± 0.1070.363 ± 0.0650.0739 ± 0.036882.2 ± 13.62.69 ± 0.4330.486 ± 0.1450.0695 ± 0.01516107 ± 26.43.21 ± 0.4440.307 ± 0.1220.0313 ± 0.00324174 ± 34.66.78 ± 1.080.474 ± 0.0460.0249 ± 0.00232257 ± 57.77.34 ± 1.300.230 ± 0.1370.0497 ± 0.014 BCF (L/Kg)

442.0

15.2

1.92

NC 060037.0 ± 15.2BDLBDL0.0294 ± 0.026249.6 ± 9.111.21 ± 0.1440.348 ± 0.0370.170 ± 0.108454.3 ± 5.821.71 ± 0.3690.396 ± 0.1420.125 ± 0.052885.8 ± 77.92.35 ± 2.050.366 ± 0.2590.161 ± 0.15716101 ± 16.33.03 ± 0.5580.324 ± 0.0960.0439 ± 0.01524157 ± 25.14.07 ± 0.6900.284 ± 0.0490.0265 ± 0.01832187 ± 23.23.10 ± 0.9340.0646 ± 0.0360.0575 ± 0.002 BCF (L/Kg)

285

8.42

0.553

NC 0120037.0 ± 15.2BDLBDL0.0294 ± 0.026259.0 ± 5.981.61 ± 0.5640.473 ± 0.1140.216 ± 0.060483.6 ± 11.32.42 ± 0.1620.742 ± 0.5290.226 ± 0.104893.9 ± 18.12.69 ± 0.4310.344 ± 0.0230.0819 ± 0.02916111 ± 10.13.21 ± 0.1530.319 ± 0.0570.0303 ± 0.00824149 ± 5.793.65 ± 0.2990.277 ± 0.0450.0657 ± 0.09732227 ± 11.22.79 ± 0.3110.0463 ± 0.0460.0601 ± 0.023 BCF (L/Kg)

342

8.92

0.290

NC Bioconcentration factors were calculated based on day 32 liver concentrations and average water concentrations measured over the entire uptake period. Average concentrations are displayed ± standard deviation. ww = wet weight. NC = not calculated. BDL = Below detection Limit. *Indicates that concentrations were below detection limits in at least one of the triplicate samples

Table 4. Average PFAS concentrations in Bluegill liver during the 32-day uptake period at 25 °CDay of ExperimentConductivity (µS/cm)PFOS Concentration (ng/g ww)PFHxS Concentration (ng/g ww)PFOA Concentration (ng/g ww)PFHxA Concentration (ng/g ww)030016.7 ± 5.020.637 ± 0.288BDL2.15 ± 2.03212.7 ± 5.521.95 ± 1.190.311 ± 0.1180.124 ± 0.202410.6 ± 5.092.43 ± 0.8360.173 ± 0.0640.0751 ± 0.096*827.2 ± 10.62.14 ± 0.6030.160 ± 0.032BDL1631.3 ± 10.61.61 ± 0.2890.088 ± 0.029BDL2472.4 ± 4.522.18 ± 0.6720.165 ± 0.171BDL3283.5 ± 33.02.23 ± 0.5490.116 ± 0.1060.0189 ± 0.009 BCF (L/Kg)

151

2.77

0.964

NC 060016.7 ± 5.020.637 ± 0.288BDL2.15 ± 2.03*26.72 ± 4.261.03 ± 0.2460.211 ± 0.0860BDL422.9 ± 9.391.46 ± 0.6500.180 ± 0.1310.168860.4 ± 41.04.38 ± 3.510.335 ± 0.3320.1931640.5 ± 16.51.70 ± 0.4120.191 ± 0.156BDL2464.3 ± 36.02.19 ± 1.020.119 ± 0.0794BDL3257.5 ± 4.141.59 ± 0.6630.0995 ± 0.0373BDL BCF (L/Kg)

121

2.71

0.879

NC 0120016.7 ± 5.020.637 ± 0.288BDL2.15 ± 2.03210.4 ± 10.71.27 ± 0.4460.179 ± 0.05171.59 ± 0.896418.5 ± 3.061.71 ± 0.4290.106 ± 0.06721.23 ± 0.344818.2 ± 7.241.16 ± 0.0920.0491 ± 0.0191.87 ± 0.3571667.0 ± 15.92.04 ± 0.6020.167 ± 0.1920.573 ± 0.4682466.0 ± 18.32.00 ± 0.7700.0923 ± 0.0750.583 ± 0.6043288.5 ± 23.61.58 ± 0.2280.0422 ± 0.0290.947 ± 0.103 BCF (L/Kg)

191

2.46

0.248

2.80 Bioconcentration factors were calculated based on day 32 liver concentrations and average water concentrations measured over the entire uptake period. Average concentrations are displayed ± standard deviation. ww = wet weight. NC = not calculated. PFOS concentrations were corrected based on contamination in control fish not exposed to PFAS. BDL = Below detection Limit. *Indicates that concentrations were below detection limits in at least one of the triplicate samples

Respirometry Study

Maximum metabolic rate, resting metabolic rate, and absolute aerobic scope in juvenile bluegill were lower in 16 °C compared to 20 and 25 °C, but there was no notable effect of conductivity on aerobic performance (Fig. 3; ANOVA temperature effects: F_RMR_ = 44.72, p-val_RMR_ < 0.001; F_AAS_ = 18.16, p-val_AAS_ < 0.001; F_MMR_ = 72.21, p-val_MMR_ < 0.001). MMR, RMR, and AAS, were higher in larger fish showing common hypoallometric mass scaling effects (scaling exponents: bRMR = 0.82, bMMR = 0.82, bAAS = 0.82; Supplemental Figure S2, Fig. 3). Only MMR was marginally affected by conductivity treatment, where the MMR was the lowest in the lowest conductivity fish (300 µS/cm, ANOVA conductivity effects: F_MMR_ = 3.128, p-val_MMR_ = 0.049; Tukey post hoc: p-val [SPC 300 vs. SPC 600] = 0.055). The factorial aerobic scope was not different between fish acclimated to any temperature and conductivity combination, suggesting that there was no aerobic constraint in any fish.

Fig. 3. Metabolic performances of juvenile bluegill in three temperatures and three ecologically relevant freshwater salinization conditions. (A) Resting metabolic rate (RMR). (B) Maximum metabolic rate (MMR). (C) Absolute Aerobic Scope (AAS = MMR - RMR). (D) Factorial Aerobic Scope (FAS = MMR/RMR). Each individual value is plotted as small, faded symbols; the predicted marginal means, and error are plotted in the dark large symbols. The letters show the difference between temperatures. The shape is conductivity. Body mass scaling exponents (b) are shown on the panel. All data were normalized to the mean size 2.5 g fish using performance-specific metabolic scaling relationships. T = Temperature, SPC = Specific Conductivity

Discussion

The overarching goal of this study was to advance understanding of the factors influencing the bioconcentration of PFAS in fish under a range of environmentally relevant conditions. In our bioconcentration experiments, bluegill were exposed to an environmentally relevant mixture of PFAS (PFOS, PFHxS, PFOA, and PFHxA; Brown et al. 2023) over a 32-d uptake and 32-d elimination period at two temperatures (20 and 25 °C) and three conductivities representing ecologically relevant levels of freshwater salinization (300, 600, and 1200 µS/cm; Moore et al. (2020). In parallel experiments, we investigated the potential influence of temperature and conductivity on fish respirometry.

In our study, calculated BCF values across all PFAS tested were generally higher in fish exposed at 20 °C relative to 25 °C, indicating reduced bioconcentration at higher temperatures following 32-d exposure. For PFHxS, this appeared to be related to faster elimination in fish exposed at higher temperatures, whereas no clear differences in elimination were observed for other PFAS tested. Similar temperature effects on the elimination were found in a dietary exposure study on rainbow trout, Oncorhynchus mykiss, where PFOS and PFHxS elimination were organ-specific, with more rapid elimination of PFOS and PFHxS from the liver at 19 °C compared to 7 °C (Vidal et al. 2019). The same study also showed lower elimination from kidneys with increasing temperature, thus highlighting that temperature mediates uptake, elimination, and bioconcentration in an organ- and compound-specific way (Vidal et al. 2019). In contrast, Wang et al. (2023) suggested that PFAS uptake/elimination rates are greater at higher temperatures in zebrafish, Danio rerio (Wang et al. 2023). However, the Wang et al. (2023) study utilized sediment-water systems, where elevated temperatures impacted sediment-water partitioning of PFAS resulting in elevated PFAS water concentrations. Therefore, differences in bioconcentration may have been related to changes in aqueous exposure rather than uptake and elimination. Similarly, a number of field-based studies have documented seasonal changes in PFAS concentrations and profiles in abiotic matrices, including surface water, sediment, and soil, relating to changes in precipitation and temperature (Brown et al. 2023; Lee et al. 2020; Pignotti et al. 2017), though concomitant seasonal variation in fish PFAS uptake has not been widely studied. Additionally, temperature changes may influence the content of fish phospholipids and proteins which in turn exert a strong influence on binding of PFAS (Dietrich et al., 2018; Sun et al. 2022). As such, alterations to plasma protein and phospholipid composition under different temperature and salinity conditions may have influenced PFAS bioconcentration. Bioconcentration (BCF values) was generally lower in fish exposed at higher conductivity levels, with some differences observed among specific PFAS. With considerable variation, several previous studies have demonstrated increased PFAS uptake or toxicity with increased salinity in fish that can tolerate a wide range of salinities (Burcham et al. 2024; Chung et al. 2024; Jeon et al. 2010; Avellán-Llaguno et al. 2020; Davis thesis 2023; Burcham et al. 2024). In black rockfish, Sebastes schlegeli, at four different salinity levels from 10–34 ppt (8–12 °C), uptake and elimination rate constants increased at higher salinities, though notable variation was observed among compounds (PFOS, PFOA, PFDA, PFUnDA) and between tissue types (liver tissue and serum) (Jeon et al. 2010). Similarly, elevated uptake of PFAS at higher salinities has been reported in marine medaka, Oryzias melastigma (PFBS, PFOS, PFOA, and PFDoA; 0–35 ppt; ~28 °C; Avellán-Llaguno et al. 2020) and killifish Fundulus heteroclitus (PFOS; 0 and ~ 25 ppt; Davis thesis 2023). Mechanistically, the increased uptake of PFOS at higher salinity has been linked to higher relative expression of organic anion transporter (OAT1) in the gill tissue, a transporter responsible for PFAS elimination across taxa (Bangma et al. 2022). Additionally, interactive effects of PFAS exposure levels and salinity can affect tissue bioconcentration; specifically, fish that were exposed to lower levels of PFOS had higher bioconcentration at low salinity, while those exposed to high levels of PFOS had higher tissue concentration at saltwater (Burcham et al. 2024). The specific mechanisms that drive these context-specific trends and contrasting results merit further study.

The decline in PFAS bioconcentration levels in liver tissue under higher conductivity in our study compared to others may be linked to relatively low PFAS exposures (e.g., < 1 µg/L), and the low range of salinities used (maximum conductivity ~ 1200 µS/cm). The broader ranges of salinity treatments in other studies likely lead to greater plasticity of physiological processes related to salinity acclimation, which may also influence PFAS uptake and elimination.

Furthermore, salting out of PFAS is known to occur at higher salinity levels, wherein aqueous solubility is reduced at higher salinities, and the enhanced fugacity of PFAS in salt water may induce molecules to move to other phases such as the gill surface for uptake (Jeon et al. 2010). The smaller range of tested salinities and lower influence of the salting out effect in the present study may have contributed to the variable and more modest effects of salinity compared with other studies. Some populations of bluegill can tolerate salinities up to 10 ppt (Peterson et al. 1993) and temperature from 18–33.5 °C, though tolerance of much lower temperatures (< 1 °C) has been recorded (Grausgruber et al. 2021; Nevada Division of Environmental Protection, 2016; Winter et al. 2018). While our study aimed to incorporate environmentally relevant temperature and salinity alterations (to mimic freshwater salinization), our results may be missing the trends under conditions nearing the functional tolerance limits of this species.

In addition to common temperatures and freshwater salinity (conductivity) ranges for bluegill, we specifically targeted environmentally relevant PFAS exposure concentrations. Often, toxicokinetic studies target concentrations over an order of magnitude higher compared to this study (e.g., Avellán-Llaguno et al. 2020; Jeon et al. 2010; Martin et al. 2003). The BCFs in the present study were substantially lower than in comparable studies using higher exposure concentrations but higher salinity (≥ 10 ppt, Jeon et al. 2010) and lower temperatures (e.g., ≤ 12 °C Martin et al. 2003; Jeon et al. 2010). This suggests potential concentration-dependence, species differences, or condition-dependence, which have been documented for PFAS (Huang et al. 2022; Lewis et al. 2022). For example, Martin et al. (2003) calculated a BCF of 100 ± 13 L/kg for PFHxS in rainbow trout, Oncorhynchus mykiss, after a 12-d uptake and 33-d depuration period following exposure to 1.4 µg/L PFHxS, compared with a maximum BCF of 15.2 L/Kg for PFHxS with 0.6 µg/L PFHxS exposure levels in the present study; since our study differs by species, concentrations, and condition, the driver of the different outcomes is unclear. However, Huang et al. (2022) exposed zebrafish, Danio rerio, to PFOS, PFHxS, and several alternatives at three concentrations but under similar temperatures (26 °C) and 500 ± 50 µS/cm conductivity conditions, finding an increase in bioconcentration of both PFOS and PFHxS at 10 µg/L relative to 1 µg/L, though a decline in bioconcentration was recorded at 100 µg/L, suggesting that concentration-dependent bioconcentration or species differences may have driven the lower bioconcentration factors in the present study. It is important to highlight that measured aqueous concentrations of PFAS were variable throughout the course of the uptake period (Table 2, S3, and S4). For example, average PFOS concentrations in the 20 °C study ranged from 0.383 ± 0.071 µg/L to 0.958 ± 0.831 µg/L based on five time points during the uptake period, which may have been driven in part by variable adsorption of PFAS to particulates. The variability in aqueous PFAS concentrations throughout the study increases uncertainty in calculated BCF values.

Significant variability was observed in liver tissue PFAS concentrations throughout the uptake and depuration periods. For example, PFOA concentrations in bluegill livers generally reached maximum concentrations in the first 10 days of exposure, followed by a decrease throughout the remainder of the uptake period despite consistent exposure. This may reflect the low exposure concentrations used (0.12 µg/L) and uncertainty in measured tissue concentrations close to analytical reporting limits rather than a true decline during the uptake period. Given the rapid elimination of PFOA observed across all treatments (Fig. 1), rapid attainment of steady state concentrations during uptake would be anticipated (Martin et al. 2003). Therefore, the slight decrease in PFOA concentrations during the uptake period observed across all treatments are likely a consequence of variability in both measured tissue concentrations and aqueous exposure concentrations rather than a true decline. Similarly, PFHxA showed no clear uptake and depuration trend across temperatures and conductivity levels, with liver concentrations close to the analytical limits of quantitation at both temperatures, despite exposure to measured concentrations ranging from 0.220–0.339 µg/L (Tables 1 and 2). Interestingly, Martin et al. (2003) exposed rainbow trout to 1.7 µg/L PFHxA but have not reported any bioaccumulation in fish liver tissue (Martin et al. 2003). It is possible that PFHxA was not accumulating unless the conditions are altered (e.g., increased conductivity). In addition, Liang et al. (2022) showed the highest elimination rates of PFHxA among a suite of many other PFAS, including those tested here. Altogether, these finding suggests lower bioconcentration of PFHxA relative to the other PFAS assessed, and rapid elimination which is consistent with previous fish PFAS bioconcentration summaries (Burkhard 2021). It remains unclear why PFHxA concentrations were elevated only at select instances in our experiment (day 0 in 25 °C fish and 25 °C fish and 1200µS/cm through the uptake and depuration phases).

Temperature, but not conductivity, influenced metabolic rates in bluegill in this study, and thus it is unlikely that alterations to metabolic rate were responsible for the interactive temperature-conductivity effects of PFAS bioaccumulation and depuration. The increase in MMR, RMR, and AS with temperature was expected and in line with reported optimal temperatures of 26–30 °C for bluegill (Lemke 1977). Additionally, the factorial aerobic scope, the measure of aerobic constraint, was unchanged (FAS > 2) across all treatments, further suggesting that all conditions were likely within a tolerable range for this commercial strain of bluegill (e.g., Eliason et al. 2022). Bluegill have been observed in brackish environments, demonstrating broad tolerance ranges for both salinity and temperature (Jones et al. 2008; Peterson et al. 1993). Studies have shown that environmental contamination (wastewater treatment plant [WWTP] effluent; similar conductivity levels ~ 1200 µS/cm to our highest treatment) can directly affect various physiological performances, including energy metabolism and energy storage in bluegill (Du et al. 2018). Combined, these studies found increased RMR, decreased survival and growth, decreased energy stores, and upregulated compensatory mechanisms to aid in sufficient oxygen supply to the tissue (e.g., increased heart mass, increased gill surface area, reduced blood oxygen affinity; Du et al. 2018, 2019). Among other contaminants, PFAS is also found in WWTP effluent (Thompson et al. 2022). The few studies that have assessed RMR in relation to PFOS exposure in fish have found effects only at concentrations significantly higher than those measured in the environment (Xia et al. 2013, 2014, 2015; Thornton 2024). For example, topmouth gudgeon, Pseudorasbora parva, exposed to 8 mg/L PFOS had significantly increased RMR relative to controls, with Thornton (2024) finding an increase in routine metabolism in juvenile red drum, Sciaenops ocellatus, exposed to 27 mg/L PFOS at 20 °C relative to controls. The interactive effects of multiple environmental conditions and PFAS exposure on fish sublethal performance metrics remain largely unexplored.

Fish with higher RMR generally have higher gill ventilation rates, which presumably leads to higher water chemical uptake as the water passes the gill lamellae (Patra et al. 2015). Similarly, toxicant depuration rate constants have been shown to increase with increasing oxygen consumption rates and metabolic rates across compounds of varying hydrophobicity (Yang et al. 2000). The present study found higher RMR and MMR in fish at 25 °C relative to 20 °C and 16 °C (Fig. 3); thus, it is possible that the increased oxygen consumption and thus ventilation rates at higher temperatures had a greater effect on contaminant depuration than uptake, leading to the reduced bioconcentration of PFOS and PFHxS at higher temperatures. Although unlikely in this study, lower dissolved oxygen concentrations in warmer water could increase contaminant uptake though increased ventilation rates. Only a few studies have examined the combined direct effects of hypoxia and PFAS (PFBS; e.g., Sun et al. 2024), but the knowledge on the indirect effects of hypoxia on PFAS uptake is limited. Further studies are needed to disentangle the relationships between toxicokinetic parameters, environmental conditions, and fish physiology.

Implications

Abiotic factors such as temperature and conductivity have the potential to influence the bioaccumulation of PFAS but field studies have not shown consistent trends in fish. Previous empirical and modeling studies have emphasized the importance of understanding physiological parameters, including respiration rate, in determining bioaccumulation of contaminants such as PFAS (Kelly et al. 2024; Sun et al. 2022; Vidal et al. 2020; Xiong and Li 2024). The present study utilized environmentally relevant PFAS exposure concentrations and realistic temperature and freshwater conductivity alterations to determine the influence of these abiotic factors and fish oxygen uptake on PFAS bioconcentration. Overall, bioconcentration factors in the liver tissue were lower in fish exposed to PFAS mixtures at higher temperatures and higher conductivity, driven largely by increased PFAS elimination. Although our findings suggest that bioconcentration of select PFAS may decrease under warming, the present study used a simplified water-only exposure that is not fully representative of exposures in the natural environment. A large contribution to total PFAS in fish tissues has been attributed to dietary exposure (Sun et al. 2022; Martin et al. 2003). Under warming, the dietary intake increases in fish to meet the increasing energy demands with temperature, and warming also can impact partitioning of PFAS between sediment and water potentially leading to great exposure from the aqueous pathway (Pignotti et al. 2017; Wang et al. 2023). Further studies across more diverse fish species and life stages (Thornton 2024thesis) will be required to gain a more nuanced understanding of how tissue PFAS bioconcentration and toxicity change due to interactive effects of fish energy metabolism, conductivity, temperature, and PFAS mixture exposure.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary Material 1