Environment International 129 (2019) 194–207

Contents lists available at ScienceDirect

Environment International

journal homepage: www.elsevier.com/locate/envint

Review article Understanding the bioavailability of pyrethroids in the aquatic environment using chemical approaches T ⁎ Zhijiang Lua,b, , Jay Ganb, Xinyi Cuic, Laura Delgado-Morenod, Kunde Line a College of Environmental and Resource Sciences, Zhejiang Provincial Key Laboratory of Agricultural Resources and Environment, Zhejiang University, Hangzhou 310058, China b Department of Environmental Sciences, University of California, Riverside, CA 92521, United States c State Key Laboratory of Pollution Control and Resource Reuse, School of the Environment, Nanjing University, Nanjing 210046, China d Estación Experimental del Zaidín, Spanish National Research Council (CSIC), Granada, Spain e The Fujian Provincial Key Laboratory for Coastal Ecology and Environmental Studies, College of the Environment and Ecology, Xiamen University, Xiamen 361102, China

ARTICLE INFO ABSTRACT

Handling Editor: Robert Letcher Pyrethroids are a class of commonly used insecticides and are ubiquitous in the aquatic environment in various Keywords: regions. Aquatic toxicity of pyrethroids was often overestimated when using conventional bulk chemical con- Pyrethroid centrations because of their strong hydrophobicity. Over the last two decades, bioavailability has been re- Bioavailability cognized and applied to refine the assessment of ecotoxicological effects of pyrethroids. This review focuses on Tenax extraction recent advances in the bioavailability of pyrethroids, specifically in the aquatic environment. We summarize the Passive sampling development of passive sampling and Tenax extraction methods for assessing the bioavailability of pyrethroids. Risk assessment Factors affecting the bioavailability of pyrethroids, including physicochemical properties of pyrethroids, and quality and quantity of organic matter, were overviewed. Various applications of bioavailability on the as- sessment of bioaccumulation and acute toxicity of pyrethroids were also discussed. The final section of this review highlights future directions of research, including development of standardized protocols for measure- ment of bioavailability, establishment of bioavailability-based toxicity benchmarks and water/sediment quality criteria, and incorporation of bioavailability into future risk assessment and management actions.

1. Introduction example, Li et al. (2017) reviewed 58 studies on the occurrence of pyrethroids in sediment, and found that the detection frequencies for Pyrethroids are a class of commonly used insecticides. Their struc- bifenthrin, cyfluthrin, lambda-cyhalothrin, cypermethrin, esfenvale- tures are analogous to naturally-occurring pyrethrins found in the rate, and permethrin were 78, 37, 49, 48, 43, and 57%, respectively, flowers of Chrysanthemum, but with increased insecticidal potency as with the maximum concentrations reaching the μg/g range (dry weight, well as environmental stability. Because of their high efficacy and low dw). mammalian toxicity, synthetic pyrethroids have gained popularity for Although pyrethroids are known to have low mammalian toxicity, both agricultural and urban insect management in recent decades as they exhibit high levels of toxicity toward aquatic invertebrates and fish some organophosphate and carbamate insecticides were gradually (Nowell et al., 2016; Weston et al., 2015). For example, the median phased out (Kuivila et al., 2012; Lao et al., 2010). Although pyrethroids effective concentration (EC50) of bifenthrin for mayfly(Hexagenia sp.) are considered to be less recalcitrant to degradation than organo- immobilization was only 15.3 ng/L (Weston et al., 2015). The threshold chlorine insecticides, they can last for months in soils because of their effect benchmark of bifenthrin in freshwater sediment for midge enhanced chemical stability compared to pyrethrins, such as stability (Chironomus spp.) was only 0.55 μg/g-OC (Nowell et al., 2016). As a against photolysis (Meyer et al., 2013). As a result of their widespread result, pyrethroids have been acknowledged as a major contributor to use and relatively long persistence, residues of pyrethroids have been sediment (Cheng et al., 2017; Holmes et al., 2008; Weston et al., 2013; found ubiquitously in the aquatic environment in many regions in the Weston et al., 2005; Yi et al., 2015) and surface water toxicity (Wolfram world (Li et al., 2017; Moschet et al., 2017; Moschet et al., 2014; et al., 2018). Schreiner et al., 2016; Van Metre et al., 2017; Wei et al., 2017). For Traditionally, the bulk or total chemical concentrations have been

⁎ Corresponding author at: College of Environmental and Resource Sciences, Zhejiang University, Hangzhou 310058, China. E-mail addresses: [email protected], [email protected] (Z. Lu). https://doi.org/10.1016/j.envint.2019.05.035 Received 29 December 2018; Received in revised form 27 April 2019; Accepted 13 May 2019 Available online 23 May 2019 0160-4120/ © 2019 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/BY-NC-ND/4.0/). .L,e al. et Lu, Z.

Table 1 Summary of the literature using passive sampling methods to obtain chemical activities of pyrethroids in the aquatic environment.

Compound Passive sampler Matrix Duration Technique Ffree (%) Reference

8 Pyrethroids SR 500 μm × 50 mm × 100 mm Freshwater sediment 7 d (in situ) Equilibrium sampling with PRC; in Xu et al., 2018 LDPE 25 μm × 50 mm × 150 mm 30 d (ex situ) situ and ex situ 8 Pyrethroids PU 381 μm × 5 mm × 50 mm River water 4 d Equilibrium sampling with PRC; in 0.27–47 Liao et al., 2017 situ 8 Pyrethroids LDPE 25 μm × 20 mm × 200 mm River water 7 d Equilibrium sampling with PRC; in 3–83 Xue et al., 2017 situ Cypermethrin, permethrin POCIS with 0.2 g Oasis HLB sorbent River water 28 d Kinetic sampling Zhang et al., 2016 9 Pyrethroids SPMD, size unknown River water 2 months Equilibrium sampling with PRC; in (Hapke et al., 2016 situ 10 Pyrethroids SR 0.5 mm × 10 cm × 30 cm River water 14 d Kinetic sampling; in situ Moschet et al., 2014 Bifenthrin, permethrin LDPE 60 μm × 2.5 cm × 98 cm Water 10 d Kinetic sampling O'Brien et al., 2012 SR 410 μm × 2.5 cm × 92 cm Fenvalerate, allethrin, SPMD 2.9 cm × 50 cm DI water 20 d Kinetic sampling Sabaliunas et al., 1998 Bifenthrin PDMS-SPME 7 μm × 200 mm DI water with 5 mg/L 40 min Kinetic sampling; online 45–65 Delgado-Moreno et al., commercial DOC 2015 Permethrin LDPE, 2.6 cm × 1 m River water 14 d Equilibrium sampling with PRC; in Anderson et al., 2014 situ Cypermethrin GF/F filter coated with ethylene vinyl acetate 7 μm × 70 mm Sea water Up to 16 d Equilibrium sampling; in situ Tucca et al., 2014 195 diameter 8 Pyrethroids PDMS-SPME 100 μm, length unknown Estuary sediment 29 d Equilibrium sampling; in situ Greenstein et al., 2014 Bifenthrin, cis- permethrin PDMS-SPME 35 μm × 10 mm Freshwater sediment 1, 2, 3, and 20 d Equilibrium sampling with and Cui et al., 2013a without PRC; ex situ 6 Pyrethroids PDMS-SPME 10 μm × 200 mm River sediment 28 d Equilibrium sampling; ex situ 1.6–5.3 Li et al., 2013 9 Pyrethroids PDMS-SPME 10 μm × 150 mm Freshwater sediment 42 d Equilibrium sampling; in situ and ex Harwood et al., 2013b situ Bifenthrin PDMS-SPME 10 μm × 400 mm Soil and freshwater sediment 28 d Equilibrium sampling; ex situ Harwood et al., 2013a Permethrin PDMS-SPME 10 μm × 10 mm Freshwater sediment 28 d Equilibrium sampling; ex situ Ding et al., 2013 5 Pyrethroids SPME with etched stainless steel River water 30 min Kinetic sampling; online Jia et al., 2010 Permethrin PDMS-SPME 10 μm × 100 mm Freshwater sediment 3, 7, 14, 28 d Equilibrium sampling; ex situ Pehkonen et al., 2010 7 Pyrethroids PDMS-SPME 30 μm × 10 or 20 mm Freshwater sediments 20 d Equilibrium sampling; ex situ 2.3–20.7 Hunter et al., 2009 Permethrin PDMS-SPME 30 μm × 10 mm Freshwater sediments 20 min Kinetic sampling; online 0.4–13.0 Hunter et al., 2008 Permethrin PDMS-SPME 10 μm × 10 mm Freshwater sediments 28 d Equilibrium sampling; ex situ 0.3, 0.4, 2.1, 5.3 You et al., 2007 Permethrin, cyfluthrin PDMS-SPME 35 μm × 30 mm Surface waters 24 h Kinetic sampling; 43–98 Yang et al., 2007 fl μ –

Bifenthrin, cy uthrin, PDMS-SPME 30 m × 20 mm Sediment porewater 24 h Kinetic sampling; 2.9 27 Xu et al., 2007 Environment International129(2019)194–207 fenpropathrin 9 Pyrethroids PDMS-SPME 30 μm × 20 mm Freshwater and marine 20 min Kinetic sampling; online 4.1 to 37; Bondarenko et al., 2007 sediment porewater 3.2 to 13.3 4 Pyrethroids PDMS-SPME 30 μm × 20 mm Freshwater and marine 15 min Kinetic sampling; online 1.7–4.4; Bondarenko et al., 2006 sediment porewater 1.1–2.9 Bifenthrin, cis- and trans- PDMS-SPME 30 μm × 20 mm Stream water and runoff 5–80 min Kinetic sampling; online 0.4–1.0; Liu et al., 2004 permethrin 10–27 Bifenthrin, cis-and trans- PDMS-SPME 30 μm × 20 mm River sediments 10 min Kinetic sampling; online Lee et al., 2003 permethrin Z. Lu, et al. Environment International 129 (2019) 194–207 used to develop toxicity benchmarks for aquatic contaminants. For se- The principles and differences of these two approaches were well de- diment, to account for the dominant role of organic carbon (OC) in scribed in Reichenberg and Mayer (2006) and Cui et al. (2013b) and are regulating phase distribution of hydrophobic contaminants, OC nor- out of the scope of this review. Both of these approaches have been used malized concentrations have also been used in establishing toxicity to characterize the bioavailability of pyrethroids in aquatic systems. thresholds and sediment quality criteria. However, pyrethroids are ex- tremely hydrophobic, with log KOW values ranging from 4.5 to 7.0 3.1. Passive sampling for measuring the chemical activity of pyrethroids (Laskowski, 2002), and log KOC values ranging from 5.08 to 6.04 (Gan et al., 2008). The use of total chemical concentration may cause over- Table 1 summarizes the studies using various passive samplers to estimated toxicity in both the water column and sediment (Mayer et al., measure the chemical activity or Cfree of pyrethroids, including semi- 2014; Yang et al., 2006c). The use of OC normalization was also pro- permeable membrane devices (SPMD), ethylene vinyl acetate coated blematic due to differences in properties of organic matter (OM) or GF/F filter, polar organic chemical integrative sampler (POCIS), film- dissolved organic matter (DOM) (Li et al., 2013; Nowell et al., 2016; type passive samplers (polyethylene devices, polyurethane (PU) film, You et al., 2008). Therefore, like other hydrophobic organic con- silicone rubber (SR) sheets), and solid phase microextraction (SPME) taminants (HOCs), such as dichlorodiphenyltrichloroethane (DDT), fibers. It should be noted that silicone rubber sheets in the reviewed polycyclic aromatic hydrocarbons (PAHs), and polychlorinated biphe- studies were based on polydimethylsiloxane (PDMS), and they were nyls (PCBs) (Alexander, 2000; Cui et al., 2013b), chemically measured abbreviated as SR in order to be distinguished from PDMS-SPME fibers. bioavailability serves as a better predictor for toxicity of pyrethroid in Sabaliunas et al. (1998) used SPMD containing triolein to mimic the water column and sediment. uptake of fenvalerate and allethrin by mussels and reported that the Recently, several reviews have been published on the bioavailability uptake rates by SPMD were about 3.5 to 5.5 times greater than those by of legacy HOCs (Booij et al., 2016; Cui et al., 2013b; Greenberg et al., mussels. Hapke et al. (2016) used SPMD to monitor various organic 2014; Knauer et al., 2017; You et al., 2011); however, so far no review contaminants including 9 pyrethroids in river water, in which 0.14 and has specifically focused on pyrethroids despite their widespread use and 0.13 ng/L of cyfluthrin was detected in two of five deployments. environmental occurrence. Pyrethroids differ from legacy HOCs in that To simplify sample cleanup, other types of passive samplers were they have comparatively shorter persistence (Lao et al., 2012), rela- later used in place of SPMD (Table 1). O'Brien et al. (2012) reported tively more rapid biotransformation (Ding et al., 2012a), and sub- that the specific sampling rates (Rs) of permethrin and bifenthrin on stantially higher acute toxicity to aquatic species. Therefore, the mea- low-density polyethylene (LDPE) and SR were a function of water ve- suring protocols and applications of bioavailability of pyrethroids can locity, and the rates can be calibrated in situ by the dissolution of be different from those of other HOCs. Here we provide an up-to-date gypsum. Moschet et al. (2014) used SR sheets to measure the con- and in-depth review on recent research advances on the measurement centrations of 10 pyrethroids in multiple rivers. However, the authors and application of bioavailability in risk assessment of pyrethroids in did not determine the passive sampler-water partition coefficients (Kpw) aquatic environments. Knowledge gaps for future research are subse- or Rs for any of the target pyrethroids. Therefore, the exact Cfree could quently outlined. not be derived. Anderson et al. (2014) developed a pesticide-mon- itoring project in the water of agricultural systems across six western 2. Literature collection and evaluation African countries, validating the robustness of LDPE samplers for

measuring Cfree of HOCs, including pyrethroids. Xue et al. (2017) Although we aimed this article to be a critical review, rather than an compared 4 polymer films, i.e., LDPE, PU, polyoxymethylene (POM), exhaustive, comprehensive systemic review, effective literature collec- and polymethylmethacrylate (PMMA), in their uptake of 8 pyrethroids tion and evaluation would ensure the quality and thoroughness of this in deionized (DI) water and showed that the uptake kinetics followed review. The online literature search was conducted using the string the order of LDPE > PMMA > PU ≈ POM. In a subsequent test, LDPE TS = (pyrethroid* OR bifenthrin OR permethrin OR cyfluthrin OR cy- was selected for sampling pyrethroids in surface water. Since the permethrin OR deltamethrin OR esfenvalerate OR fenpropathrin OR equilibrium of pyrethroids between LDPE and water was not reached cyhalothrin) AND TS = (bioavailability OR bioaccessibility OR passive within 30 d, preloading of the LDPE film with stable isotope-labeled sampl* OR Tenax OR SPME or LDPE) in Web of Science, in which TS bifenthrin and cis-permethrin as performance reference compounds means topic, and then similar strings were used for literature search in (PRCs) was introduced to derive the equilibrium concentrations in the Scopus and PubMed. The search was limited to English-language peer- LDPE film (Xue et al., 2017). The principle of PRC calibration may be reviewed research articles. All collected references with abstracts were found in other publications (Fernandez et al., 2009). Liao et al. (2017) imported to Endnote X8 (Clarivate Analytics, Boston, MA). After re- further developed film-type passive samplers for pesticide monitoring moving duplicated references, the title and abstract of each reference in surface water including 8 pyrethroids. The authors compared six (total of 331) were examined to assess if the publication met the fol- types of thin films including polyethersulfone, polycarbonate, POM, SR, lowing inclusion criteria: (1) studies were carried out in aquatic sys- PU, and polyvinyl chloride and reported that SR and POM both showed tems; (2) bioavailability was measured by chemical approaches; (3) if satisfactory sorption of pyrethroids while PU was the best. Xu et al. passive sampling was used, procedures, for example, heating, were not (2018) extended Xue et al. (2017)’s study to freshwater sediments and used to interrupt the natural equilibrium. Afterward, full texts of reported that LDPE and SR showed higher affinity than PMMA, POM, screened publications (total of 83) were obtained and further evaluated and PU films. Both LDPE and SR were further tested for in situ appli- by the above inclusion criteria and the quality of the study: (1) suffi- cation at several freshwater sediment sites in Southern California, and cient description on methods to reproduce the study; (2) sample du- the Cfree values derived from these two types of passive samplers were plicates; (3) at least basic statistics, such as t-test and standard errors. ideally regressed with the slope being 1.00 ± 0.02 and R2 being 0.99.

Finally, 72 research articles were selected for this review. Moreover, the Cfree values derived from PRC-corrected 7-d in situ sampling and 30-d ex situ equilibrium sampling were also linearly 3. Chemical approaches to estimate pyrethroid bioavailability correlated, demonstrating the feasibility of using PRC-correction on film type passive samplers to allow in situ measurements with a short Generally, there are two chemical approaches to measure the deployment time. bioavailability of HOCs. One aims to measure the freely dissolved Although several film-type passive samplers have been tested for concentration (Cfree) or the chemical activity; while the other aims to measuring chemical activities of pyrethroids, SPME fibers have been estimate the bioaccessibility, i.e., the labile fraction of HOCs that can used in the majority of studies to date (Table 1). It should be noted that readily desorb from the solid phase (Reichenberg and Mayer, 2006). not all SPME methods can be classified as an approach to measure

196 Z. Lu, et al. Environment International 129 (2019) 194–207 chemical activity. For example, Barrionuevo and Lancas (2000, 2002) steady state for several days under mixing conditions; therefore, non- and Casas et al. (2006) used PDMS-SPME fibers to measure con- equilibrium or kinetic sampling were commonly used for the injector- centrations of pyrethroid in enriched river water. However, heating type SPME (Table 1). The influence of matrix effects on the accuracy of

(from 30 to 90 °C) was used to facilitate the sampling process, which Cfree analysis in the kinetic sampling needs to be carefully evaluated, could result in disturbed phase distribution. Similarly, in the head-space since Rs values could be altered by media constituents, for example, SPME method, the equilibrium was also disturbed by either heating or humic acid (Cui et al., 2013b), although such effects have not been microwave irradiation (Casas et al., 2006; Li et al., 2009; You and Lydy, reported for pyrethroids yet. The disposable SPME offers several other 2006). Therefore, these methods can not be considered as methods for advantages, including much lower cost and greater flexibility allowing measuring chemical activities. However, comparisons of different SPME kinetic and equilibrium sampling, and matching the typical duration of fibers in these studies could still provide intrinsic information on pyr- bioassays. For example, disposable fibers may be placed in the same test ethroid-fiber interactions. Li et al. (2009) reported that the PDMS fiber vessel used for bioassays for toxicity or bioaccumulation measurement exhibited the highest sorption capacity compared to the other five (Harwood et al., 2012a; Harwood et al., 2012b). In such an application, commercial fibers. Casas et al. (2006) similarly observed that poly- SPME fibers and test organisms are exposed to the same conditions, and acrylate and PDMS fibers had the highest efficacy among four candidate accumulation of pyrethroids into fibers could provide mechanistic in- fibers. Jia et al. (2010) used etched stainless steel fibers as SPME sights into factors affecting the bioaccumulation and toxicity in the test samplers to measure residues of pyrethroids in soil and water. The species (Harwood et al., 2012a, 2012b). enrichment factors of five pyrethroids on this sampler were about 3 to Both film and fiber types of passive samplers have been deployed for 10 times of those on commercial PDMS fibers. In subsequent studies, in situ field monitoring while only the fiber type has been used in the ex however, PDMS fibers were most used because of PDMS's outstanding situ measurement (Table 1). However, no study has yet compared the enrichment capability toward pyrethroids (Table 1). advantages and disadvantages of these two types of samplers. Gen- Gan and colleagues introduced injector-type PDMS fibers to mea- erally, because of their larger sampler volume, film type samplers have sure the chemical activities of several pyrethroids in surface water and a higher sensitivity, which is essential for field applications. For ex- sediment (Bondarenko et al., 2006; Bondarenko et al., 2007; Delgado- ample, the volume of the LDPE sampler was 100 μL Moreno et al., 2015; Lee et al., 2003; Liu et al., 2004)(Table 1). In this (20 mm × 200 mm × 25 μm) (Xue et al., 2017) while the volume of the application, injector-type PDMS fibers were immersed into agitated SPME fiber was about 2.67 μL (15 cm × 10 μm) (Harwood et al., water or sediment samples for a fixed time interval. The whole fibers 2013b). The cost of using injector type SPME is higher than that for were then injected into a gas chromatography (GC) inlet, and the disposable SPME and film type passive samples, and most disposable analytes enriched on the fiber were thermally desorbed for analysis. SPME fibers used glass as the supporting material, which is usually too While manual sampling and injection were used in initial studies fragile to be inserted into sediment unless protected by additional metal (Bondarenko et al., 2006; Bondarenko et al., 2007; Lee et al., 2003; Liu housings. For example, disposable SPME fibers were housed into copper et al., 2004), the same method was automated using an autosampler tubes for in situ sampling of pyrethroids (Greenstein et al., 2014) and capable of sampling at a fixed depth and for a fixed time interval, other HOCs (Witt et al., 2013) in sediment. sample mixing, injection, and thermal desorption (Delgado-Moreno Some other types of passive samplers were also used to measure the et al., 2015). Disposable SPME fibers were also introduced to measure Cfree of pyrethroids. Tucca et al. (2014) used GF/F filters coated with Cfree of pyrethroids in water and sediment (Table 1). In this method, ethylene vinyl acetate as passive samplers to measure concentrations of SPME fibers were cut into desired lengths and then placed in direct cypermethrin in salmon farms in Chile. Polar organic chemical in- contact with matrices (water or sediment). After a fixed time, the fibers tegrative samplers which are usually used for hydrophilic organic were recovered and extracted with organic solvents. Performance re- contaminants were also used for regulatory monitoring of nine pesti- ference compounds were also introduced for SPME (Cui et al., 2013a), cides including cypermethrin and permethrin in River Ugie, Scotland identical Cfree values of pyrethroids were obtained within much shorter (Zhang et al., 2016). time by PRC-SPME compared to the SPME method without PRC. Injector-type and disposable SPME offer different advantages. As all ffi the pyrethroids sorbed on the whole fiber were desorbed into the inlet, 3.2. Passive sampler-water partition coe cients for pyrethroids high analytical sensitivity of injector-type SMPE was consistently achieved. With the use of autosampler, sampling conditions were pro- When passive samplers are used for equilibrium sampling, Cfree is fi grammed and kept constant, leading to highly reproducible measure- obtained by dividing the concentration on ber (Cfiber) over Kpw. ments (Delgado-Moreno et al., 2015). However, the accumulation of Therefore, Kpw values are an essential parameter for calculation of Cfree pyrethroids from the sample media onto the PDMS fiber did not reach a (Booij et al., 2016; Lydy et al., 2014). Table 2 summarizes Kpw values of several commonly used pyrethroids on various passive samplers. It is

Table 2 Summary of passive sampler-water partition coefficients of pyrethroids in the logarithmic form.

Compound log KOW 500 μm 381 μm 25 μm 100 μm PDMS in 100 μm 30 μm PDMS 10 μm PDMS 7 μm PDMS 30 μm PDMS (Laskowski, SR (Xu PU (Liao LDPE seawater (Lao PDMS (Hunter et al., (You et al., (Bondarenko (Bondarenko 2002) et al., et al., (Xue et al., 2012) (Lao et al., 2009) 2007) et al., 2007) et al., 2007) 2018) 2017) et al., 2012) 2017)

Bifenthrin 6.4 5.10 4.23 5.18 5.87 ± 0.10 5.69 5.97 ± 0.2 4.26 ± 0.10 3.79 ± 0.09 cis-Permethrin 4.33 5.16 6.43 ± 0.10 6.24 5.46 ± 0.3 4.57 ± 0.06 3.94 ± 0.08 trans-Permethrin 5.16 6.33 ± 0.10 6.14 5.37 ± 0.3 4.55 ± 0.01 3.96 ± 0.04 Permethrin 6.9 4.17 5.59 ± 0.04 Cyfluthrin 5.97 4.54 3.89 5.06 5.93 ± 0.12 5.75 5.65 ± 0.1 4.25 ± 0.10 3.88 ± 0.19 Cypermethrin 6.54 4.37 3.95 5.05 5.99 ± 0.12 5.79 6.26 ± 0.1 4.32 ± 0.08 3.98 ± 0.14 Deltamethrin 4.53 4.70 5.72 4.92 5.89 ± 0.13 5.70 4.01 ± 0.10 3.58 ± 0.14 Esfenvalerate 5.62 4.73 4.48 5.12 6.33 ± 0.13 6.15 5.80 ± 0.03 4.05 ± 0.14 3.52 ± 0.12 Fenpropathrin 6.0 4.87 5.02 5.12 5.73 ± 0.10 5.56 5.51 ± 0.1 4.52 ± 0.03 4.07 ± 0.01 λ-Cyhalothrin 7.0 4.63 4.72 5.15 6.12 ± 0.11 5.92 4.21 ± 0.08 3.76 ± 0.25

197 Z. Lu, et al. Environment International 129 (2019) 194–207

Fig. 1. Relationships between passive sampler-water and octanol-water partition coefficients of pyrethroids. (A) Film type passive samplers. Dots, 500 μmSR(Xu et al., 2018); triangles, 381 μmPU(Liao et al., 2017); squares, 25 μm LDPE (Xue et al., 2017). (B) PDMS-SPME fibers. Dots, 25 μm PDMS (Lao et al., 2012); triangles, 30 μm PDMS (Hunter et al., 2009); squares, 7 μm PDMS (Bondarenko et al., 2007); diamonds 30 μm PDMS (Bondarenko et al., 2007).

obvious that the Kpw values among various materials were different, equilibrium was reached, albeit significant hydrolysis was observed for due to different sorption affinities of pyrethroids to these materials. It is cyfluthrin, cypermethrin, and esfenvalerate. Lao et al. (2012) also re- surprising that the Kpw values for a given pyrethroid on PDMS fibers ported that for pyrethroids like cyfluthrin, the apparent equilibrium among different studies also varied significantly. For example, the log could not be reached due to strong hydrolysis, and a two-compartment

Kpw values of bifenthrin on PDMS fibers ranged from 3.79 to 5.97 model had to be used to calculate the corresponding Kpw values. In the (Table 2). Generally, the values followed the order: 100 μm PDMS (Lao two-compartment model, due to the loss via hydrolysis, steady-state et al., 2012)>30μm PDMS (Hunter et al., 2009)>7μm PDMS cannot be reached, and the Cfiber and Cw follow Eqs. (3) and (4): (Bondarenko et al., 2007)>30μm PDMS (Bondarenko et al., 2007). It Cw,0 −−βt αt should be noted that the PDMS fibers used by Lao et al. (2012) and Cwt, = [(kβe21−−− ) ( kαe ) ] αβ− (3) Bondarenko et al. (2007) were purchased from Supelo (Bellefonte, PA, USA) while those used by Hunter et al. (2009) were from Polymicro kC1,0w −−βt αt ⎤ Technologies (Phoenix, AZ, USA). Although the variation in K might Cfiber, t = []ee− pw αβ− ⎥ be caused by the between-manufacturer variability and differences in ⎦ (4) ff coating thicknesses (as a result of di erent manufacturing processes), where α and β are hybrid-rate constants for uptake by and elimination previous studies on other HOCs have demonstrated that contributions from the fiber, respectively, k1 and k2 are the uptake and elimination from these two factors were negligible (Booij et al., 2016; DiFilippo and rate constants, respectively. Parameters k1, α, and β can be obtained by Eganhouse, 2010). Moreover, all three studies were carried out at the fitting Eq. (4). Parameter k2 can be obtained by: room temperature; therefore, the temperature should not be an im- portant contributor to the variation. Therefore, methodologies used to αβ k2 = b (5) determine Kpw values must be the major source of the observed dis- crepancy (DiFilippo and Eganhouse, 2010). Since partitioning equili- where b is the aqueous phase hydrolysis-rate constant which can be brium could not be reached within hours, K cannot be simply derived pw obtained by an independent time-series hydrolysis experiment. The Kpw by dividing Cfiber over Cw. Instead, a one-compartment model such as values are calculated as: that in Bondarenko et al. (2007) was used to estimate Kpw under non- k1 steady conditions. In the one-compartment model, loss processes such Kpw = k2 (6) as hydrolysis are not included, and Cfiber is fitted to a one-compartment rise to maximum equation: Nevertheless, the Kpw values of cyfluthrin, cypermethrin, and es- −kt fenvalerate by Lao et al. (2012) and Hunter et al. (2009) were within CCfiber,, t=×− fiber eq (1 e) (1) 0.4 logarithmic unit. This may be attributed to the interlaboratory fi where Cfiber,t and Cfiber,eq are concentrations on the ber at time t and at variation, which is often within 0.2 to 0.5 logarithmic unit (Booij et al., equilibrium, respectively, and k is the rate constant. The Kpw values are 2016). Xue et al. (2017) also used the one-compartment model to ob- calculated as: tain partition coefficients for 9 pyrethroids between LDPE film and C C water. Overall, the Kpw values of pyrethroids need to be further in- fiber, eq fiber, t −kt Kpw ==/(1 −e ) vestigated, especially intercomparisons between methods, laboratories, Cweq, Cwt, (2) and sampler types are necessary. where Cw,t and Cw,eq are the aqueous concentrations at time t and at DiFilippo and Eganhouse (2010) illustrated that log Kpw values on equilibrium, respectively. Although the same model was used by PDMS fibers and log Kow values of PAHs and DDTs were linearly cor- 2 Hunter et al. (2009) to estimate the Kpw of bifenthrin, a larger value was related with R values > 0.90. However, no such regression existed for obtained, likely because a much longer incubation time (33 d) was pyrethroids on film (Fig. 1a) or fiber-type (Fig. 1b) passive samplers. It used. Therefore, the one-compartment model may underestimate Kpw should be noted that in DiFilippo and Eganhouse (2010), data from values when the status is far from equilibrium. In Bondarenko et al. Bondarenko et al. (2007) were mistakenly cited, i.e., the Kpw values in (2007), the phenomenon that Kpw values on the 7-μm fiber were much DiFilippo and Eganhouse (2010) were exactly 3 orders of magnitude greater than those on the 30-μm fiber further supported this conclusion. lower than those originally reported by Bondarenko et al. (2007). The

Hunter et al. (2009) calculated the Kpw values for other pyrethroids updated regressions between log KOW and log Kpw of pyrethroids are except for bifenthrin using the ratio of Cfiber over Cw after the apparent given in Fig. 1. The regression for LDPE was statistically significant

198 Z. Lu, et al. Environment International 129 (2019) 194–207

Table 3 Summary of the literature using Tenax resin to obtain bioaccessibility of pyrethroids in the aquatic environment.

Compound Tenax mass Sediment mass Water volume Tenax: OC TOC (%) Duration F (%)a Reference (g) (g) (dw) (mL) Ratio

Bifenthrin 0.5 3 45 4.3–64 0.26–3.89 24 h Hartz et al., 2018 Cypermethrin 0.5 5 (wet) 45 0.82–7.99 Sequential Zhang et al., 2018 7 Pyrethroids 0.510–1.65 3 40 5–60 1.70–4.24 24 h Nutile et al., 2017b 7 Pyrethroids 0.084–3.64 3 40 5–60 0.56, 2.04 24 h Nutile et al., 2017a 7 Pyrethroids 0.5 3 40 0.56, 2.04 Sequential Nutile et al., 2017a 6 Pyrethroids 1 3 40 2.5–59.5 0.56–13.3 24 h 17–55 Li et al., 2013 9 Pyrethroids 0.5 2 40 7.4–20.8 1.2–3.4 24 h Harwood et al., 2013b Bifenthrin 0.5 40 6 h, 24 h Harwood et al., 2013a 10 Pyrethroids 2 2 (wet) 40 5.8, 2.0 Sequential 10–20; 15–40 Feo et al., 2013 Permethrin 0.5 2.5 45 12.5 1.6 24 h Du et al., 2013 4 Pyrethroids 0.1 0.1 31 3.18 Sequential (Fojut and Young, 2011) Permethrin 0.3 11 100 6.8 0.4 Sequential 11.4, 12.4 Cui et al., 2010 Bifenthrin, permethrin 0.05 0.2 14 19 1.31 Sequential 18–40; 32–60 You et al., 2009 7 Pyrethroids 0.5 3 45 3.4–50.5 0.33–4.92 6 h You et al., 2008 7 Pyrethroids 0.05 1 20 3.3 1.5 Sequential 11–14 Yang et al., 2008 4 Pyrethroids 0.15 2 25 1.5–5.2 1.44–5.03 Sequential 11.4–45.4 Xu et al., 2008 Permethrin 0.05 0.2 14 3.18, 19.1 1.31, 7.85 Sequential 15.9–51.1 You et al., 2007

a Frap or the ratio of C6hor C24 h and the total concentration.

(R2 = 0.65, p = 0.008), but the slope of regression was very small ratio > 5 for the single-point Tenax extraction. In a later study, the (0.084 ± 0.023, p = 0.008); therefore, differences among pyrethroids authors further emphasized that the Tenax: OC ratio > 5 would be might be smaller than the measurement errors (Fig. 1a). Beside of that, adequate to ensure the quality of extraction (Nutile et al., 2017b). only regressions with one film type (381 μmPUbyLiao et al., 2017) fi μ and one PDMS ber (7 m PDMS by Bondarenko et al., 2007) passive 3.4. Isotope dilution method samplers were marginally significant with R2 both of 0.46 and p both of fi 0.045, respectively, while all others were not statistically signi cant. It An isotope dilution method (IDM) was adopted to measure the ff should be noted that the di erences of Kpw values among pyrethroids bioavailability of HOCs including pyrethroids by Gan and his colleagues were between 0.26 and 0.89 logarithmic unit except those reported by (Delgado-Moreno and Gan, 2013; Delgado-Moreno et al., 2015; Jia ff Liao et al. (2017) (Table 2). These di erences were much smaller than et al., 2014; Wang et al., 2018). The detailed principles can be found in those for PAHs (DiFilippo and Eganhouse, 2010) and thus could be Delgado-Moreno and Gan (2013). Briefly, a small amount of an isotopic easily masked by measurement errors. analog (isotopomer, isotopic isomer) is added into a partition system, such as sediment-water and dissolved organic matter (DOM)-water, and 3.3. Tenax extraction for estimating bioaccessibility of pyrethroids the isotopomer can exchange rapidly with the native HOCs within the exchangeable pool but not with those in the irreversible sorbed pool. Although several methods have been developed to estimate the The most important assumption of this method is that isotopomers bioaccessibility of HOCs, such as mild solvent and cyclodextrin ex- behave identically to the native (non-labeled) HOCs in the sorption tractions (Cui et al., 2013b), sequential and single-point Tenax extrac- process (Delgado-Moreno and Gan, 2013). For the IDM method, the tions are still the most popular methods (Table 3). In the sequential most critical variable is the mixing time which should be long enough extraction method, Tenax beads are removed, and fresh beads are to ensure equilibrium within the exchangeable pool but not too long to ff added at incremental time intervals (Feo et al., 2013; Nutile et al., allow the isotopomer to di use into the irreversible bound fraction. 2017a; Xu et al., 2008; Yang et al., 2008; You et al., 2009; You et al., Delgado-Moreno and Gan (2013) chose 48 h for pyrene and bifenthrin 2007). The desorption profile was fitted to a three-compartment model in a sediment-water system, while in a later study, 16 h was used for developed by Cornelissen et al. (1997). The rapid desorption fraction these compounds in a DOM-water system (Delgado-Moreno et al., 2015). These time intervals were determined through preliminary ex- Frap is considered to be the bioaccessible fraction. The sequential ex- traction, however, is time-consuming (14 d or even 100 d) and tedious periments, and thus may introduce additional uncertainties in the (usually > 10 extractions in total). Therefore, single-point Tenax ex- measurement. traction (just one extraction after 6 h or 24 h) has been introduced as an alternative (Table 3). You et al. (2008) found that the 6-h Tenax ex- 3.5. Comparisons of passive sampling, Tenax exaction, and the IDM method tracted concentration (C6h) underestimated the toxicity of pyrethroids to Hyalella azteca, although it was better than the total concentration, There were several studies comparing different measurement and then suggested that 24-h extraction might be a better choice for risk methods for bioavailability of pyrethroids. You et al. (2006) reported assessment. Yang et al. (2008) documented that in the sequential ex- that both single-point Tenax extraction and SPME methods provided traction, the desorption fraction after 24 h was statistically equal to Frap matrix- and chemical-independent estimations of bioaccumulation of while the fraction after 6 h was only about half of the Frap for pyre- four HOCs (including permethrin) in two sediments. Delgado-Moreno throids. Harwood et al. (2012b) reported that 24-h Tenax extracted and Gan (2013) reported that the accessible concentration (Ca) of bi- concentrations (C24h) of permethrin and bifenthrin in sediment were fenthrin measured by the IDM method was about twice as the Frap 1.65 ± 0.25 times of C6h. Harwood et al. (2013a) concluded that the measured by the Tenax extraction and was also about the sum of Frap relationship between C6h and C24h in three sediments with various and Fs (the slow desorption fraction by Tenax). Harwood et al. (2012b) spiked concentrations was log C24h = 0.952 ( ± 0.035) log C6h + 0.390 tried to regress Cfiber with C6h or C24h and found that the regression was 2 2 ( ± 0.093), R = 0.96. Nutile et al. (2017a) concluded that the ration of compound specific. For example, the R of the regressions with Cfiber C24h and Crap (the rapid desorption concentration in the sequential and C24h for permethrin and bifenthrin were 0.99 and 0.97, respec- extraction) was about 1.46 and aging slightly affected the ratio with a tively; however combining these two compounds led to a decrease of R2 daily factor of −0.0027. The authors also recommended a Tenax: OC to 0.72 (Harwood et al., 2012b). This relationship was further extended

199 Z. Lu, et al. Environment International 129 (2019) 194–207

from one sediment to three sediments in a follow-up study (Harwood later study, Cfree of permethrin in sediment and biota-sediment accu- et al., 2013a), revealing that the regression was not sediment-specific; mulation factors (BASF) of Lumbriculus variegatus were not significantly for example, the relationship for bifenthrin in three sediments was log altered by the addition of charcoal and soot (Pehkonen et al., 2010). 2 C24h = 0.971 ( ± 0.037) log Cfiber + 0.284 ( ± 0.010), R = 0.97. However, a recent study reported that 3% addition of three carbonac- Harwood et al. (2013b) compared three methods (field-deployed SPME, eous materials (activated carbon, charcoal, and biochar) reduced Cfree laboratory-exposed SPME, and 24-h Tenax extraction) on the bioa- of four pyrethroids in marine sediments by 78–85% (Li et al., 2018a). vailability measurement of 9 pyrethroids in 6 sediments, and concluded The differences between Yang et al. (2009) and Li et al. (2018a) may be that the 10-μm PDMS fiber was not as sensitive as Tenax extraction and attributed to the different properties of the sorbents used in these stu- that the field-deployed SPME was less sensitive than the laboratory- dies. exposed SPME. The poor sensitivity of SPME fibers was due to their The contact time or aging can be another important factor affecting much smaller volume. The sensitivity of SPME could be greatly im- the bioavailability of pyrethroids, in both field and laboratory amended proved if injector-type fibers were used (Hunter et al., 2009). Linear sediments (Bondarenko et al., 2006; Ding et al., 2013; Hartz et al., regressions of Frap and Cfiber of 7 pyrethroids in black carbon-amended 2018; Nutile et al., 2017a; Xu et al., 2008; You et al., 2009). Because of sediments was also reported (Yang et al., 2008). Overall, Cfree, Frap, and the heterogeneity of organic matter, it took months for pyrethroids to Ca values may be considered to be significantly linearly related, al- reach equilibrium between sediment and water phases (Bondarenko though they stand for different aspects of bioavailability, i.e., chemical et al., 2006). After 30-d incubation, the freely dissolved fractions (Ffree) activity and bioaccessibility, respectively. of pyrethroids in sediment were significantly reduced compared to

those after 9-d incubation. For example, Ffree decreased from 4.3 to 4. Factors affecting bioavailability of pyrethroid 16.3% to 1.7–4.4% and from 1.6 to 4.1% to 1.1–2.9% in a freshwater sediment and a marine sediment, respectively. Xu et al. (2008) quan-

Sorption is the inherent process controlling the bioavailability of tified the decreasing kinetics of Frap of four pyrethroids in two sedi- pyrethroids, and therefore factors that can impact sorption can also ments. The decrease in Frap followed a first-order decay model, with affect bioavailability. These factors include, but are not limited to, half-lives for four pyrethroids ranging from two to three months, which physicochemical properties of pyrethroids, the quality and quantity of were shorter than the biodegradation half-lives of the corresponding (dissolved) organic matter, particle sizes of sediment, the content of pyrethroids. Similar results were reported by Fojut and Young (2011). suspended solids, aging, and salinity. The physicochemical properties The desorption rate constant of the slow compartment of permethrin in are the intrinsic factor affecting the bioavailability of pyrethroids. sediment aged for 28 d was significantly lower compared to those only Hunter et al. (2009) spiked 7 pyrethroids in three freshwater sediments aged for 1 d and 7 d. Hartz et al. (2018) reported the effect of sample at the same level and observed that the Cfree values of these pyrethroids holding time (the length of time a sample is stored after collection and in all sediments under both agitated and nonagitated conditions gen- prior to analysis) on C24h and toxicity of bifenthrin in field-collected erally followed the same order: cypermethrin < cyfluthrin, esfenvale- contaminated sediments: In sediments collected in 2016 when there rate < fenpropathrin, permethrin, bifenthrin. Feo et al. (2013) illu- was more fresh deposition, the C24h values showed first-order decay strated that the desorption kinetics of 10 pyrethroids in two sediments with half-lives ranging from 3 to 45 d of holding, whereas in samples by Tenax extraction followed the same order: lambda-cyhalothrin < collected in 2017, linear loss of up to 14% during 180 d was observed. bifenthrin < deltamethrin < tetramethrin < fluvalinate < There have been limited studies on effects of other factors, including permethrin < esfenvalerate/fenvalerate < cypermethrin < cyfluthrin. particle sizes of sediment, suspended solids, and salinity. Zhang et al.

Several studies focused on the effects of the quality and quantity of (2018) reported unexpected results that Frap of cypermethrin decreased (dissolved) organic matters on pyrethroid bioavailability (Delgado- with increasing particle sizes of sediment, which may be attributed to

Moreno et al., 2010; Delgado-Moreno et al., 2015; Feo et al., 2013; the adsorption of fine particles to Tenax beads. However, Cfree values in Fojut and Young, 2011; Yang et al., 2007; Yang et al., 2006b; You et al., porewater of fine sediment fractions as determined by PDMS-SPME 2009; You et al., 2007; Xue et al., 2017). It was apparent that the OC or were still lower than those of coarse fractions The only studies on the DOC content could decrease the chemical activity of pyrethroids since impact of suspended solids on pyrethroid bioavailability and conse-

Cfree accounted for the truly dissolved fraction, not the OC-sorbed quently aquatic toxicity were Yang et al. (2006a, 2006c). The authors fraction. For example, when adding just 2 mg/L Aldrich humic acid into concluded that the relative contribution of suspended solids over that of water, Cfree of pyrethroids decreased by 19–49% (Xue et al., 2017). You dissolved organic matter depended on the OC content and the texture of et al. (2009) reported that the bioaccessibility of pyrethroids was in- source sediments. The effect of suspended solids was dominant when hibited by high OC contents, showing that not all the OC-sorbed frac- sediment was sandy and the organic matter content was < 1%. Because tion was bioaccessible, validating that sorption of HOCs to organic of the salting-out effect, salinity may increase the sorption of HOCs to matter was a dual-compartment process (Gustafsson and Gschwend, sediment and organic matter and thus, decreases the bioavailability of

1997). Fojut and Young (2011) also reported that both the rapid des- HOCs. Xue et al. (2017) reported that Cfree of pyrethroids in seawater orption fractions and the rapid desorption rate constants of pyrethroids with 3.5 g/L sea salt was about half of that in freshwater. Also, salinity decreased as OC contents of suspended solids increased. The properties may increase Kpw of pyrethroids. The Kpw values of pyrethroids on of (dissolved) organic matters are also critical (Table 4). Yang et al. PDMS fiber were 0.18 logarithmic unit higher in seawater than those in

(2007) reported that the carboxylic acid content of OC was linearly freshwater (Lao et al., 2012). The impact of salinity on Kpw can be es- related to KOC values of permethrin and cyfluthrin whereas no such timated by the Setschenow equation (DiFilippo and Eganhouse, 2010). relationship was found for the phenolic acid content. Delgado-Moreno et al. (2015) documented that DI water amended with Aldrich humic 5. Applications of chemically measured bioavailability of acid had lower Cfree, Frap, and exchangeable pools than those treated Pyrethroids with organic matter of other origins such as Suwannee River humic acid and fulvic acid, and Pahokee peat humic acid. Black carbon (e.g., soot, 5.1. Estimate bioaccumulation of pyrethroids in benthic organisms char) is a specific form of OC, and numerous studies have shown that it can strongly adsorb HOCs, such as PCBs and PAHs (Cornelissen et al., One of the most important applications of bioavailability is to es- 2005; Gustafsson and Gschwend, 1997). However, for pyrethroids, the timate bioaccumulation of pyrethroids in benthic organisms. sorption capacity of charcoal was found to be moderate and comparable Pyrethroids are acutely toxic to aquatic invertebrates at very low levels, to soil or sediment OC; for example, when adding 1% of charcoal to and test organisms cannot survive or maintain ordinary behaviors in sediment, Cfiber only decreased by 5.7 to 9.1% (Yang et al., 2009). In a some heavily contaminated sediments. Therefore, it is necessary to

200 .L,e al. et Lu, Z.

Table 4 − Summary of (dissolved) organic carbon-water partition coefficients (105 Lkg 1) of pyrethroids determined by passive samplers.

Matrix Bifenthrin cis-Permethrin trans-Permethrin Cyfluthrin Fenpropathrin λ-Cyhalothrin Cypermethrin Esfenvalerate Deltamethrin Reference

Soil DOC 0.48 ± 0.02 0.69 ± 0.01 0.68 ± 0.04 0.56 ± 0.03 Delgado-Moreno et al., Potting mix DOC 0.81 ± 0.05 2.39 ± 0.16 1.88 ± 0.13 1.88 ± 0.13 2010 Compost DOC 2.50 ± 0.07 3.95 ± 0.12 3.17 ± 0.09 2.35 ± 0.11 Sediment OC 1.54–6.20a Hunter et al., 2008 Sediment DOC 3.5–38.4a Surface water DOC 0.16–0.79a 0.29-1.36 Yang et al., 2007 Sediment OC 3.8–9.8 3.3–4.4 3.7–4.7 Xu et al., 2007 Sediment DOC 1.7–3.5 3.5–4.4 0.8–1.2 Sediment DOC 1.80 ± 0.15 1.31 ± 0.08 0.69 ± 0.09 1.90 ± 0.44 0.33 ± 0.02 1.70 ± 0.11 1.41 ± 0.11 3.39 ± 0.34 5.31 ± 099 Bondarenko et al., 2007 201 Marine sediment DOC 6.24 ± 1.19 6.07 ± 0.69 4.91 ± 0.58 5.98 ± 0.98 3.35 ± 0.75 8.55 ± 1.26 3.85 ± 0.73 13.5 ± 2.09 14.3 ± 2.44 Lakewater, pond water, compost 0.47 ± 0.28; 1.12 ± 0.14; Yang et al., 2006b extract DOC 0.310 ± 0.07a Riverwater DOC 0.68–2.16 0.54–1.26a Yang et al., 2006a Riverwater POC 2.35–3.34 2.73–5.74a Sediment OC 7.0 ± 0.2 18.0 ± 5.4 8.8 ± 3.4 31.4 ± 8.6 Bondarenko et al., 2006 Sediment DOC 27.4 ± 4.9 14.5 ± 3.4 31.5 ± 5.5 31.0 ± 7.4 Marine sediment OC 2.7 ± 0.5 3.5 ± 0.9 2.2 ± 0.5 6.1 ± 1.9 Marine sediment DOC 22.6 ± 3.4 13.5 ± 3.2 20.6 ± 2.8 28.4 ± 7.1 Streamwater and runoff DOC 2.9,2.5 1.6,1.6 1.6,1.8 Liu et al., 2004 River sediment DOCb 1.9, 2.8, 2.1 11.2, 3.1, 5.8 3.8, 2.2, 2.9 Lee et al., 2003 River sediment OCb 95, 45, 55 108, 53, 69 53, 46, 31

a Reported as permethrin. b Converted from KDOM and KOM by multiplying by 1.72. Environment International129(2019)194–207 Z. Lu, et al. Environment International 129 (2019) 194–207

estimate bioaccumulation of pyrethroids measured by chemical ap- proaches rather than bioassays directly. Several studies have successfully estimated body residues (BR) of pyrethroids in aquatic invertebrates using bioavailability derived by PDMS-SPME, Tenax extraction, and IDM (Table 5). In the early studies,

Cfiber was often used in the regression with BR (Ding et al., 2012a; Harwood et al., 2012b; Hunter et al., 2008). For example, Hunter et al. (2008) established such a regression model using concentrations of fi Hunter et al., 2008 Harwood et al., 2012b Delgado-Moreno et al., 2015 Delgado-Moreno and Gan, 2013 You et al., 2009 Ding et al., 2012a Ding et al., 2012a Ding et al., 2013 Reference permethrin in Chironomus tentans in several eld-collected sediments (Table 5). The same approach was also adopted by Ding et al. (2012a, 2 R 0.88 0.73 0.91 0.54 0.91 0.99 2013) to estimate residues of permethrin and bifenthrin in Hyalella azteca and Chironomus dilutus in reconstituted water and sediment. The

advantage of this approach is that no Kpw values are needed, and thus, uncertainties from Kpw values are excluded. However, models from

1.83( ± different studies cannot be compared with each other due to the dif- − - 1.12( ± 0.07) 0.98 - 0.74( ± 0.09) 0.96 fi - 0.58( ± 0.10) 0.97 - (1.25 ± 0.01) 0.99

0.26( ± 0.06) 0.75 ference in experimental details (e.g., ber thickness and length, in- ber fi ber ber – + 0.45( ± 0.08) 0.64 ber fi ber fi C fi fi cubation time). Recently, Li et al. (2018a) found that C in marine C free C C ber ber C fi fi C C sediments was linearly correlated with lipid-normalized concentrations of pyrethroids in clams, even after the sediment was amended with

various types and levels of carbonaceous materials, suggesting Cfree as a good predictor for tissue residues and also as an effective tool for evaluating remediation effectiveness. Accessible concentration mea- sured by IDM (Delgado-Moreno and Gan, 2013; Delgado-Moreno et al., + 0.27( ± 0.25)

rap 2015), C6h and C24h measured by the single-point Tenax extraction 4.63 0.88 F

− (Harwood et al., 2012b), and Frap measured by sequential Tenax ex- a C 0.9536 0.72 traction (You et al., 2009) have also been used to estimate BR in various – + 11.6 a

C invertebrates. However, it was reported that BSAF of permethrin in ber fi

C Chironomus tentans from the slow desorption pool was 60.7% of that

= 1.23( ± 0.096) × log BR + 0.150( ± 0.16) from rapid desorption pool in the non-amended sediment, whereas in = 0.948 ( ± 0.039) × log BR + 0.630 ( ± 0.069) = 0.954 ( ± 0.044) × log BR + 0.819 ( ± 0.071) 3.0 × ber

fi 6h 24h charcoal-amended sediments, the value decreased to 25.3% (Cui et al., = C C C

org 2010). Therefore, not only the Frap but also the slow desorption fraction Equation C log log log (lipid normalized BR-4d = 1.07 ( ± 0.01) × log log BSAF = 1.39 ( ± 0.53) × log BR = 2.41 × log log (lipid normalized BR-10d = 0.88( ± 0.10) × log log (lipid normalized BR-4d = 0.94( ± 0.10) × log log (lipid normalized BR-10 d = 1.27( ± 0.08) × log 0.10) log (lipid normalized BR-10 d = 1.08( ± 0.12) × log log (lipid normalized BR) = 0.83( ± 0.13) × log BR = 0.0248 × contributed to the bioaccumulation of pyrethroids. If there was a universal regression model that fits all the pyre- throids, organisms, and matrices, the application of bioavailability on the estimation of bioaccumulation could be greatly promoted. You et al.

(2009) reported that the regression of BASF and Frap was significantly 2

s log (lipid normalized BR) = 0.50( ± 0.11) × log improved when only permethrin was included, i.e., R increased from 0.54 with both permethrin and bifenthrin to 0.98 with permethrin only. Harwood et al. (2012b) reported that the regression between Tenax extracted concentrations and BR was statistically identical for perme- thrin and bifenthrin, although the regression including both compounds Chironomus dilutus Chironomus dilutu Lumbriculus variegatus, Hexagenia sp. Lumbriculus variegatus Chironomus tentans Chironomus tentans Chironomus dilutus Chironomus dilutus Hyalella azteca Hyalella azteca Daphnia magna was more scattered compared with those including only one single compound. Moreover, the regressions were also identical for two spe- cies (Lumbriculus variegatus and Hexagenia sp.). However, Ding et al. (2013) observed that in a sediment, Chironomus dilutus accumulated more permethrin than Hyalella azteca; thus two distinct regression models were derived (Table 5). The difference may be caused by the Technique Organism SPME SPME 24-h Tenax 6-h Tenax different biotransformation kinetics in these two invertebrates (Ding et al., 2013) since in an earlier study, they found that biotransformation of permethrin was more efficient in Hyalella azteca than in Chironomus dilutus (Ding et al., 2012a). In reconstituted water, the regression

models between BR and Cfiber of permethrin and bifenthrin were spe- cies-specific, incubation time-specific, and also compound-specific (Ding et al., 2012a)(Table 5). For example, Hyalella azteca accumulated more permethrin than Chironomus dilutus after 10 d; however, the trend was the opposite after 4-d incubation. The authors also documented that Chironomus dilutus metabolized bifenthrin and permethrin, while Hyalella azteca only transformed permethrin. This phenomenon further Matrix Reconstituted water Freshwater sediments suggested that biotransformation of pyrethroids should be considered when establishing prediction models of BR using bioavailability. For a given pyrethroid-invertebrate combination, the regression model seems to be consistent among field-collected sediments, which was validated with permethrin in Chironomus tentans (Hunter et al., 2008), bifenthrin in Chironomus tentans (Delgado-Moreno and Gan, 2013) and Daphnia magna (Delgado-Moreno et al., 2015). Compound Permethrin Permethrin Freshwater sediments SPME Permethrin, bifenthrin Freshwater sediments SPME Bifenthrin, permethrin Freshwater sediment Sequential Tenax Permethrin Bifenthrin Freshwater sediments IDM Bifenthrin Reconstituted water SPME Bifenthrin DI water with 5 mg/L commercial DOM IDM fi Table 5 Summary of the literature predicting bioaccumulation by pyrethroid bioavailability. A few studies showed that there was no signi cant correlation

202 Z. Lu, et al. Environment International 129 (2019) 194–207

Table 6 Summary of bioavailability-based aquatic toxicity benchmarks of pyrethroids.

Compound Matrix Invertebrate Benchmarks Value Reference

Bifenthrin Freshwater sediment Chironomus tentans LC50–10 d Cfree at 48–53 ng/L Xu et al., 2007

Hyalella azteca LC50–10 d Cfiber at 0.149–0.163 μg/mL Harwood et al., 2013a

Cfree at 0.540–0.594 ng/L

C6h at 60.096–0.206 μg/g-OC

C24h at 0.204–0.368 μg/g-OC

EC50–10 d (Immobilization) Cfiber at 0.110–0.124 μg/mL

Cfree at 0.400–0.458 ng/L

C6h at 0.576–2.74 μg/g-OC

C24h at 1.49–4.57 μg/g-OC

Chironomus dilutus LC50–10 d Cfiber at 0.905–2.44 μg/mL

Cfree at 3.28–8.84 ng/L

C6h at 0.576–2.74 μg/g-OC

C24h at 1.49–4.57 μg/g-OC

EC50–10 d (Immobilization) Cfiber at 0.411–1.37 μg/mL

Cfree at 1.49–4.97 ng/L

C6h at 0.225–1.39 μg/g-OC

C24h at 0.626–2.75 μg/g-OC

Reconstituted water Chironomus dilutus LC50–4d Cfiber at 8.3 μg/mL Ding et al., 2012b

EC50–4 d (Immobilization) Cfiber at 5.2 μg/mL

LC50–10 d Cfiber at 2.8 μg/mL

EC50–10 d (Immobilization) Cfiber at 1.4 μg/mL

Hyalella azteca LC50–4d Cfiber at 0.06 μg/mL

LC50–10 d Cfiber at 0.06 μg/mL

Permethrin Freshwater Daphnia magna LC50–48 h Cfiber at 2.26 μg/mL after 48 h Harwood et al., 2012a

Cfiber at 3.76 μg/mL in equilibrium

Freshwater sediment Chironomus dilutus LC50–10 d Cfiber at 1.45 μg/mL Ding et al., 2013

Hyalella azteca LC50–10 d Cfiber at 0.86 μg/mL

Reconstituted water Chironomus dilutus LC50–4d Cfiber at 19.6 μg/mL Ding et al., 2012b

EC50–4 d (Immobilization) Cfiber at 9.3 μg/mL

LC50–10 d Cfiber at 9.2 μg/mL

EC50–10 d (Immobilization) Cfiber at 7.8 μg/mL

Hyalella azteca LC50–4d Cfiber at 1.9 μg/mL

LC50–10 d Cfiber at 2.3 μg/mL

Freshwater sediments Chironomus dilutus LC50–20 d C24h at 1.13 ± 0.898 μg/g-OC Du et al., 2013 a LC50–58 d C24h at 0.6 μg/g-OC a EC50-Growth C24h at 0.5 μg/g-OC a EC50-Emergence C24h at 0.4 μg/g-OC

EC50-Reproduction C24h at 0.187 ± 0.071 μg/g-OC

Cyfluthrin Freshwater sediment Chironomus tentans LC50–10 d Cfree at 8.7–8.9 μg/L Xu et al., 2007

Fenpropathrin Freshwater sediment Chironomus tentans LC50–10 d Cfree at 6.1–6.8 μg/L Xu et al., 2007

Tefluthrin Reconstituted water Chironomus dilutus LC50–4d Cfiber at 160 μg/mL Ding et al., 2012b

EC50–4 d (Immobilization) Cfiber at 98.0 μg/mL

LC50–10 d Cfiber at 107 μg/mL

EC50–10 d (Immobilization) Cfiber at 52.6 μg/mL

Hyalella azteca LC50–4d Cfiber at 59.7 μg/mL

LC50–10 d Cfiber at 10.7 μg/mL

a Not exact values. Values were estimated from Fig. 5B of Du et al. (2013). between BR and bioavailability of pyrethroids. Harwood et al. (2012b) normalized toxicity benchmarks (for example, LC50 values normalized reported that the BR of permethrin in Daphnia magna did not regress to OC concentrations) (Li et al., 2013; Yi et al., 2015; You et al., 2008). with the Cfiber in laboratory reconstituted moderately hard water and in Bioavailability-based toxicity benchmarks showed significant ad- natural lake water. The authors attributed this to suspended solids in vantages over traditional approaches. Xu et al. (2007) established water or extensive biotransformation of permethrin by Daphnia magna bioavailability-based aquatic toxicity benchmarks for bifenthrin, cy- since only about 30% of permethrin remained in the organism after 48- fluthrin, and fenpropathrin in freshwater sediment by PDMS-SPME and h bioassays (Harwood et al., 2012b). Nutile et al. (2017b) also reported a Chironomus tentans sediment toxicity test (Table 6). The coefficients of that C24h of four pyrethroids did not correlate with BR values in Lum- variation (CV) for bioavailability (Cfree)-based LC50 were 0.9–13.0%. briculus variegatus in laboratory-spiked sediment. The authors suspected These values were much smaller than CVs of the other concentration- that the spiked concentration was high enough to cause toxic responses based LC50 values, which were 33.2–73.9%, 5.0–41.0%, 78.8–73.8%, to Lumbriculus variegatus, and a steady-state was not reached. and 3.7–41% for the bulk sediment, the OC-normalized bulk sediment, the porewater, and the DOC-normalized concentrations based values, respectively. Therefore, bioavailability based LC50 values were matrix- 5.2. Prediction of aquatic toxicity of pyrethroids independent and were more suitable for benchmarking aquatic toxicity

of pyrethroids. Harwood et al. (2013b) used the C24h of bifenthrin in six Bioavailability can be utilized to predict the acute and chronic field-collected sediments to establish dose-response curves of the mor- toxicities of pyrethroids to aquatic invertebrates. Two approaches have tality of Hyalella azteca, Chironomus dilutus, and Hexagenia sp. Surpris- been attempted: One is establishing dose-response curves based on ingly, the dose-response curves were not species-sensitive. Lydy and bioavailability and then obtaining the bioavailability-based toxicity colleagues considered other endpoints besides lethality to developed benchmarks (Table 6)(Ding et al., 2012b; Ding et al., 2013; Du et al., additional bioavailability-based toxicity benchmarks for pyrethroids in 2013; Harwood et al., 2012a; Harwood et al., 2013a; Xu et al., 2007), sediment and water using Chironomus dilutus, Hyalella azteca, and and the other is to combine bioavailability and conventional OC-

203 Z. Lu, et al. Environment International 129 (2019) 194–207

Daphnia magna (Table 6). For example, Ding et al. (2012b) introduced certain PAHs (Flores-Cervantes et al., 2009). In contrast, KDOC values immobilization as an additional endpoint for bioavailability-based can be easily derived using passive samplers since passive sampling toxicity test. There are few bioavailability-based chronic toxicity selectively detects Cfree which is used for calculating the true KDOC. benchmarks for pyrethroids. Du et al. (2013) established only chronic Using injector-type SPME, Gan et al. (2008) obtained KDOC values for benchmarks covering survival, growth, emergence, and reproduction of most pyrethroids and in a wide range of sediments. Chironomus dilutus through a whole life-cycle toxicity test and a 24-h Tenax bioaccessibility assay. (Table 6). 6. Knowledge gaps and future directions Conventional OC-normalized toxicity benchmarks were sometimes used as alternatives since the bioavailability-based toxicity benchmarks The ultimate goals of studying the bioavailability of pyrethroids are were available only for a few pyrethroid-invertebrate combinations to establish bioavailability-based water and sediment criteria and to (Table 6). In this approach, bioavailable toxic units (TUs) are calculated incorporate bioavailability into risk assessment and management ac- as the ratio of bioavailable concentrations (e.g., Cfree, C6h) and the OC- tions. However, research to date is still far away from these goals. normalized toxicity benchmarks (Li et al., 2013; Yi et al., 2015; You Below we identify some knowledge gaps and potential directions to et al., 2008). You et al. (2008) reported that when using C6h instead of further this line of research. the total concentration, the regression of TUs and sediment toxicity was improved from R2 of 0.24 to 0.60. However, the authors also noted that 6.1. Standardizing protocols for measuring the bioavailability of pyrethroids the bioavailable C6h-based TUs likely underestimated the toxicity in some sediment samples. Li et al. (2013) developed bioavailable TUs Several methods have been developed to measure the bioavail- using the C24h and Cfree of six pyrethroids and mortality of Hyalella ability of pyrethroids (Tables 1 and 3). However, no standardized azteca in field-collected urban river sediments. The authors compared protocol has been formalized, which limits the application of bioa- the predicted mortalities using both bioavailable and conventional TUs vailability in the regulatory context. The Tenax extraction methods are and found that the conventional TUs overestimated the toxicity of more consistent among different studies than the passive sampling pesticide mixtures in urban sediments in China. By comparing bioa- methods, and the sequential and 24-h single-point extraction methods vailable TUs for various pesticides, the authors also determined that the are more popular than the 6-h single point extraction method (Table 3). major contributor to sediment toxicity was bifenthrin. They further For the PDMS-SPME methods, significant variations in the fiber di- established a dose-response curve using Hyalella azteca mortality and mensions and deployment time existed, and PRCs were not consistently bioavailable Tus of bifenthrin. Bioavailable TUs were also used to used to account for the non-equilibrium sampling conditions (Table 1). identify the causative toxicants in urban sediment in Guangzhou, China Moreover, the Kpw values for PDMS fibers varied by more than two (Yi et al., 2015). The authors found that the low bioavailability of HOCs orders of magnitude (Table 2). The errors in Kpw values were believed limited their toxicity and that metals, pyrethroids, and fipronil, instead to be the most important contributor to the overall uncertainties of the of PAHs and PCBs, were major contributors to the sediment toxicity. In Cfree measurement for HOCs (Booij et al., 2016; Lydy et al., 2014). For a recent study, XAD-based bioaccessibility together with high- the IDM method and other passive samplers, to date, only a few studies throughput midge toxicity testing reveal that pyrethroids contributed to have been conducted. Moreover, most of the research to date was 17–35% of toxicity in four sediments in Guangzhou, China; whereas conducted in the laboratory, while very few studies measured the in situ without considering bioavailability, their contribution was dramatically bioavailability of pyrethroids. Therefore, future research needs to overestimated (56–1625%) (Li et al., 2019). This approach, i.e.,effect- standardize the sampling and analytical protocols preferably with ac- directed analysis with bioavailability could be used to identify the key curate Kpw values and the inclusion of PRCs. Inter-laboratory compar- toxicants and quantify their contributions (Greenstein et al., 2014; Li isons on the standard protocols would also help to validate and promote et al., 2018b; Qi et al., 2017; You and Li, 2017). Therefore, the bioa- their use. vailability of pyrethroids may not only be used to predict the toxicity of pyrethroids but also to understand the potential causes of the toxicity in 6.2. Developing bioavailability based toxicity benchmarks aquatic environments. Bioavailability based toxicity benchmarks should be further devel- 5.3. Understand the distribution of pyrethroids in aquatic environments oped. First, current bioavailability-based aquatic toxicity benchmarks are limited to a few pyrethroid-invertebrate combinations, and most Passive sampling techniques (SPME, LDPE) can be used to char- studies only considered acute lethality as the endpoint (Table 6), which acterize the partitioning of pyrethroids between water and sediment, limits the application of bioavailability in risk assessment. More pyre- suspended solids, and/or DOM. For example, Liu et al. (2004) used throid-organism combinations need to be investigated in the future, PDMS-SPME to understand the distribution of bifenthrin and perme- especially for chronic toxicity. Second, hydrolysis and sorption to thrin in runoff and in stream water. The authors found that in stream glassware were found to be significant for several pyrethroids (Lao water about 97% of bifenthrin and permethrin were bound to sus- et al., 2012), while a stable concentration is required to ensure the pended solids and 2.3–2.8% to DOM, while only < 1% was truly dis- accuracy of threshold values such as LC50s. Therefore, new techniques, solved. Accurate determination of Kd, KOC, and KDOC values of pyre- for example, passive dosing (Adolfsson-Erici et al., 2012), should be throids using passive samplers (Table 4) would also help to improve employed to maintain constant exposure concentrations. Third, bio- understanding of their distribution in the aquatic environment. When transformation of pyrethroids in tested organisms was frequently re- using traditional methods for measuring Kd and KOC values, the effect of ported (Ding et al., 2012a; Harwood et al., 2012b). It is possible that DOM is neglected, because the measured aqueous concentration in fact some toxic effects are caused by metabolites of pyrethroids, such as 3- includes both Cfree and the fraction sorbed to DOM. Thus, Kd and KOC phenoxybenzyl alcohol and 3-phenoxybenzoic acid. Therefore, effects values determined using the conventional approach are likely under- of these metabolites need to be understood. estimated. For example, Lee et al. (2003) found that Kd values of per- methrin measured by the SPME method were 1.6–5.4 and 4.6–22.7 6.3. Improving bioaccumulation prediction times higher than those obtained by the conventional liquid-liquid ex- traction in creek sediments and in nursery runoff sediments, respec- Similar knowledge gaps exist when predicting the bioaccumulation tively. Measuring KDOC was technically challenging; traditional of pyrethroids using bioavailability in aquatic organisms. First, this methods have considerable limitations. For example, the fluorescence approach has only been applied to several invertebrates. Second, the quenching method was limited to compounds that fluoresce such as toxicity of pyrethroid may alter the behavior of test organisms and thus

204 Z. Lu, et al. Environment International 129 (2019) 194–207 affect bioaccumulation. Therefore, these effects need to be avoided or Forsberg, N.D., Haigh, T.A., Sower, G.J., 2014. Passive sampling devices enable ca- compensated. Third, biotransformation of pyrethroids in test organisms pacity building and characterization of bioavailable pesticide along the Niger, Senegal and Bani Rivers of Africa. Philos. Trans. R. Soc. B 369 (1639). would lead to biases on bioaccumulation estimation by bioavailability, Barrionuevo, W.R., Lancas, F.M., 2000. Solid-phase microextraction of pyrethroid pesti- which should also be considered. Finally, as discussed in Section 5.1, cides from water at low and sub-ppt levels at different temperatures. J. High Resolut. current prediction equations (Table 5) are just based on a few pyre- Chromatogr. 23, 485–488. Barrionuevo, W.R., Lancas, F.M., 2002. Comparison of liquid-liquid extraction (LLE), throids while a universal model is still lacking. It will be helpful if solid-phase extraction (SPE), and solid-phase microextraction (SPME) for pyrethroid quantitative structure-activity relationship models are incorporated. pesticides analysis from enriched river water. Bull. Environ. Contam. Toxicol. 69, Attention should be paid to the calculation, collection, and selection of 123–128. sensitive structural parameters. To improve the model accuracy, it is Bondarenko, S., Putt, A., Kavanaugh, S., Poletika, N., Gan, J.Y., 2006. Time dependence of phase distribution of pyrethroid insecticides in sediment. Environ. Toxicol. Chem. also recommended to conduct well-controlled microcosm studies in the 25, 3148–3154. laboratory prior to the model development and field validation. Bondarenko, S., Spurlock, F., Gan, J., 2007. Analysis of pyrethroids in sediment pore water by solid-phase microextraction. Environ. Toxicol. Chem. 26, 2587–2593. Booij, K., Robinson, C.D., Burgess, R.M., Mayer, P., Roberts, C.A., Ahrens, L., Allan, I.J., 6.4. Incorporating bioavailability into management actions Brant, J., Jones, L., Kraus, U.R., Larsen, M.M., Lepom, P., Petersen, J., Pröfrock, D., Roose, P., Schäfer, S., Smedes, F., Tixier, C., Vorkamp, K., Whitehouse, P., 2016. Since bioavailability provides a better measurement of aquatic Passive sampling in regulatory chemical monitoring of nonpolar organic compounds in the aquatic environment. Environ. Sci. Technol. 50, 3–17. toxicity than the bulk concentration, it should be used in water quality Budd, R., O'Geen, A., Goh, K.S., Bondarenko, S., Gan, J., 2011. Removal mechanisms and monitoring in connection to pyrethroid contamination before a reg- fate of insecticides in constructed wetlands. Chemosphere 83, 1581–1587. ulatory or management action is taken. For example, as mentioned in Casas, V., Llompart, M., Garcia-Jares, C., Cela, R., Dagnac, T., 2006. Multivariate opti- fl ff mization of the factors in uencing the solid-phase microextraction of pyrethroid Section 5.3, most pyrethroids in surface runo were adsorbed to sus- pesticides in water. J. Chromatogr. A 1124, 148–156. pended solids while very few were freely dissolved (Liu et al., 2004). Cheng, F., Li, H.Z., Qi, H.X., Han, Q., You, J., 2017. Contribution of pyrethroids in large Therefore, in order to reduce the loading of pyrethroids via surface urban rivers to sediment toxicity assessed with benthic invertebrates Chironomus ff dilutus: a case study in South China. Environ. Toxicol. Chem. 36, 3367–3375. runo , the best management practice should be diminishing the level of Cornelissen, G., vanNoort, P.C.M., Govers, H.A.J., 1997. Desorption kinetics of chlor- suspended solids via reducing soil erosion and sediment resuspension. obenzenes, polycyclic aromatic hydrocarbons, and polychlorinated biphenyls: sedi- The removal of suspended solids was demonstrated to be highly effec- ment extraction with Tenax® and effects of contact time and solute hydrophobicity. – tive in reducing the levels of various pyrethroids in constructed wet- Environ. Toxicol. Chem. 16, 1351 1357. Cornelissen, G., Gustafsson, O., Bucheli, T.D., Jonker, M.T.O., Koelmans, A.A., Van Noort, lands (Budd et al., 2011). Similarly, sedimentation ponds and vegeta- P.C.M., 2005. Extensive sorption of organic compounds to black carbon, coal, and tive filters were found to be hyperefficient at minimizing pyrethroid kerogen in sediments and soils: mechanisms and consequences for distribution, – loads in the runoff water discharged from a large commercial nursery bioaccumulation, and biodegradation. Environ. Sci. Technol. 39, 6881 6895. Cui, X.Y., Hunter, W., Yang, Y., Chen, Y.X., Gan, J., 2010. Bioavailability of sorbed (Lao et al., 2008). phenanthrene and permethrin in sediments to Chironomus tentans. Aquat. Toxicol. 98, 83–90. 7. Conclusions Cui, X.Y., Bao, L.J., Gan, J., 2013a. Solid-phase microextraction (SPME) with stable iso- tope calibration for measuring bioavailability of hydrophobic organic contaminants. Environ. Sci. Technol. 47, 9833–9840. Bioavailability allows a better estimation of the toxicity of pyre- Cui, X.Y., Mayer, P., Gan, J., 2013b. Methods to assess bioavailability of hydrophobic throids in aquatic environments than the bulk concentration or OC organic contaminants: principles, operations, and limitations. Environ. Pollut. 172, 223–234. normalized concentration. Several chemical approaches including se- Delgado-Moreno, L., Gan, J., 2013. A stable isotope dilution method for measuring quential and single-point Tenax extractions, PDMS-SPME, LDPE passive bioavailability of organic contaminants. Environ. Pollut. 176, 171–177. sampling, and IDM have been developed to measure the bioavailability Delgado-Moreno, L., Wu, L., Gan, J., 2010. Effect of dissolved organic carbon on sorption of pyrethroids to sediments. Environ. Sci. Technol. 44, 8473–8478. of pyrethroids in aquatic matrices. From these studies, there is a good Delgado-Moreno, L., Wu, L.S., Gan, J., 2015. Application of isotope dilution method for understanding on the effects of, the quality and quantity of (dissolved) measuring bioavailability of organic contaminants sorbed to dissolved organic matter organic matter, the content of suspended solids, aging, and salinity on (DOM). Aquat. Toxicol. 165, 129–135. ffi pyrethroid bioavailability. Moreover, bioavailability has been utilized DiFilippo, E.L., Eganhouse, R.P., 2010. Assessment of PDMS-water partition coe cients: implications for passive environmental sampling of hydrophobic organic compounds. to estimate the BR and toxicity of pyrethroids in aquatic invertebrates, Environ. Sci. Technol. 44, 6917–6925. identify the causative contaminant, and elucidate the phase distribution Ding, Y.P., Landrum, P.F., You, J., Harwood, A.D., Lydy, M.J., 2012a. Use of solid phase fi of pyrethroids in the aquatic environment. However, further research is microextraction to estimate toxicity: relating ber concentrations to body residues- part II. Environ. Toxicol. Chem. 31, 2168–2174. still required to reach the ultimate goals of using bioavailability in risk Ding, Y.P., Landrum, P.F., You, J., Harwood, A.D., Lydy, M.J., 2012b. Use of solid phase assessment and management. Future directions include, but are not microextraction to estimate toxicity: relating fiber concentrations to toxicity-part I. limited to, developing standardized protocols, establishing more bioa- Environ. Toxicol. Chem. 31, 2159–2167. Ding, Y.P., Landrum, P.F., You, J., Lydy, M.J., 2013. Assessing bioavailability and toxicity vailability-based acute and chronic toxicity benchmarks, investigating of permethrin and DDT in sediment using matrix solid phase microextraction. bioavailability-based bioaccumulation in the food web, and in- Ecotoxicology 22, 109–117. corporating bioavailability into pyrethroid management practices. Du, J., Pang, J.X., You, J., 2013. Bioavailability-based chronic toxicity measurements of permethrin to Chironomus dilutus. Environ. Toxicol. Chem. 32, 1403–1411. Feo, M.L., Corcellas, C., Barata, C., Ginebreda, A., Eljarrat, E., Barcelo, D., 2013. Organic Acknowledgement carbon content effects on bioavailability of pyrethroid insecticides and validation of solid phase extraction with poly (2,6-diphenyl-p-phenylene oxide) polymer by fi Daphnia magna toxicity tests. Sci. Total Environ. 442, 497–502. This study was nancially supported by the National Key Research Fernandez, L.A., Harvey, C.F., Gschwend, P.M., 2009. Using performance reference and Development Program of China (2018YFC1800303) and the compounds in polyethylene passive samplers to deduce sediment porewater con- Fundamental Research Funds for the Central Universities centrations for numerous target chemicals. Environ. Sci. Technol. 43, 8888–8894. (2019QNA6007). Flores-Cervantes, D.X., Reddy, C.M., Gschwend, P.M., 2009. Inferring black carbon con- centrations in particulate organic matter by observing pyrene fluorescence losses. Environ. Sci. Technol. 43, 4864–4870. References Fojut, T.L., Young, T.M., 2011. Desorption of pyrethroids from suspended solids. Environ. Toxicol. Chem. 30, 1760–1766. Gan, J., Bondarenko, S., Spurlock, F., 2008. Persistence and phase distribution in sedi- fl Adolfsson-Erici, M., Åkerman, G., Jahnke, A., Mayer, P., McLachlan, M.S., 2012. A ow- ment. In: Synthetic Pyrethroids. 991. American Chemical Society, pp. 203–222. through passive dosing system for continuously supplying aqueous solutions of hy- Greenberg, M.S., Chapman, P.M., Allan, I.J., Anderson, K.A., Apitz, S.E., Beegan, C., et al., drophobic chemicals to bioconcentration and aquatic toxicity tests. Chemosphere 86, 2014. Passive sampling methods for contaminated sediments: risk assessment and – 593 599. management. Integr. Environ. Assess. Manag. 10, 224–236. Alexander, M., 2000. Aging, bioavailability, and overestimation of risk from environ- Greenstein, D.J., Bay, S.M., Young, D.L., Asato, S., Maruya, K.A., Lao, W.J., 2014. The use – mental pollutants. Environ. Sci. Technol. 34, 4259 4265. of sediment toxicity identification evaluation methods to evaluate clean up targets in Anderson, K.A., Seck, D., Hobbie, K.A., Traore, A.N., McCartney, M.A., Ndaye, A., an urban estuary. Integr. Environ. Assess. Manag. 10, 260–268.

205 Z. Lu, et al. Environment International 129 (2019) 194–207

Gustafsson, O., Gschwend, P.M., 1997. Soot as a strong partition medium for polycyclic 4702–4708. aromatic hydrocarbons in aquatic systems. ACS Symp. Ser. 671, 365–381. Moschet, C., Vermeirssen, E.L.M., Seiz, R., Pfefferli, H., Hollender, J., 2014. Picogram per Hapke, W.B., Morace, J.L., Nilsen, E.B., Alvarez, D.A., Masterson, K., 2016. Year-round liter detections of pyrethroids and organophosphates in surface waters using passive monitoring of contaminants in Neal and Rogers Creeks, Hood River Basin, Oregon, sampling. Water Res. 66, 411–422. 2011-12, and assessment of risks to salmonids. PLoS One 11. Moschet, C., Lew, B.M., Hasenbein, S., Anumol, T., Young, T.M., 2017. LC- and GC-QTOF- Hartz, K.E.H., Sinche, F.L., Nutile, S.A., Fung, C.Y., Moran, P.W., Van Metre, P.C., et al., MS as complementary tools for a comprehensive micropollutant analysis in aquatic 2018. Effect of sample holding time on bioaccessibility and sediment ecotoxicological systems. Environ. Sci. Technol. 51, 1553–1561. assessments. Environ. Pollut. 242, 2078–2087. Nowell, L.H., Norman, J.E., Ingersoll, C.G., Moran, P.W., 2016. Development and appli- Harwood, A.D., Bunch, A.R., Flickinger, D.L., You, J., Lydy, M.J., 2012a. Predicting the cation of freshwater sediment-toxicity benchmarks for currently used pesticides. Sci. toxicity of permethrin to Daphnia magna in water using SPME fibers. Arch. Environ. Total Environ. 550, 835–850. Contam. Toxicol. 62, 438–444. Nutile, S.A., Harwood, A.D., Sinche, F.L., Hartz, K.E., Landrum, P.F., Lydy, M.J., 2017a. Harwood, A.D., Landrum, P.F., Lydy, M.J., 2012b. Can SPME fiber and Tenax methods Methodological and environmental impacts on bioaccessibility estimates provided by predict the bioavailability of biotransformed insecticides? Environ. Sci. Technol. 46, single-point tenax extractions. Arch. Environ. Contam. Toxicol. 72, 612–621. 2413–2419. Nutile, S.A., Harwood, A.D., Sinche, F.L., Hartz, K.E.H., Landrum, P.F., Lydy, M.J., 2017b. Harwood, A.D., Landrum, P.F., Lydy, M.J., 2013a. Bioavailability-based toxicity end- The robustness of single-point Tenax extractions of pyrethroids: effects of the Tenax points of bifenthrin for Hyalella azteca and Chironomus dilutus. Chemosphere 90, to organic carbon mass ratio on exposure estimates. Chemosphere 171, 308–317. 1117–1122. O'Brien, D., Komarova, T., Mueller, J.F., 2012. Determination of deployment specific Harwood, A.D., Landrum, P.F., Weston, D.P., Lydy, M.J., 2013b. Using SPME fibers and chemical uptake rates for SPMD and PDMS using a passive flow monitor. Mar. Pollut. Tenax to predict the bioavailability of pyrethroids and chlorpyrifos in field sediments. Bull. 64 (5), 1005–1011. Environ. Pollut. 173, 47–51. Pehkonen, S., You, J., Akkanen, J., Kukkonen, J.V.K., Lydy, M.J., 2010. Influence of black Holmes, R.W., Anderson, B.S., Phillips, B.M., Hunt, J.W., Crane, D.B., Mekebri, A., et al., carbon and chemical planarity on bioavailability of sediment-associated con- 2008. Statewide investigation of the role of pyrethroid pesticides in sediment toxicity taminants. Environ. Toxicol. Chem. 29, 1976–1983. in California's urhan waterways. Environ. Sci. Technol. 42, 7003–7009. Qi, H., Li, H., Wei, Y., Mehler, W.T., Zeng, E.Y., You, J., 2017. Effect-directed analysis of Hunter, W., Xu, Y.P., Spurlock, F., Gan, J., 2008. Using disposable polydimethylsiloxane toxicants in sediment with combined passive dosing and in vivo toxicity testing. fibers to assess the bioavailability of permethrin in sediment. Environ. Toxicol. Chem. Environ. Sci. Technol. 51 (11), 6414–6421. 27, 568–575. Reichenberg, F., Mayer, P., 2006. Two complementary sides of bioavailability: accessi- Hunter, W., Yang, Y., Reichenberg, F., Mayer, P., Gan, J.Y., 2009. Measuring pyrethroids bility and chemical activity of organic contaminants in sediments and soils. Environ. in sediment pore water using matrix-solid phase microextraction. Environ. Toxicol. Toxicol. Chem. 25, 1239–1245. Chem. 28, 36–43. Sabaliunas, D., Lazutka, J., Sabaliuniene, I., Sodergren, A., 1998. Use of semipermeable Jia, C.H., Zhu, X.D., Zhao, E.C., Yu, P.Z., He, M., Chen, L., 2010. Application of SPME membrane devices for studying effects of organic pollutants: comparison of pesticide based on a stainless steel wire for the determination of pyrethroid insecticide residues uptake by semipermeable membrane devices and mussels. Environ. Toxicol. Chem. in water and soil. Chromatographia 72, 1219–1223. 17, 1815–1824. Jia, F., Bao, L.J., Crago, J., Schlenk, D., Gan, J., 2014. Use of isotope dilution method to Schreiner, V.C., Szocs, E., Bhowmik, A.K., Vijver, M.G., Schafer, R.B., 2016. Pesticide predict bioavailability of organic pollutants in historically contaminated sediments. mixtures in streams of several European countries and the USA. Sci. Total Environ. Environ. Sci. Technol. 48, 7966–7973. 573, 680–689. Knauer, K., Homazava, N., Junghans, M., Werner, I., 2017. The influence of particles on Tucca, F., Moya, H., Barra, R., 2014. Ethylene vinyl acetate polymer as a tool for passive bioavailability and toxicity of pesticides in surface water. Integr. Environ. Assess. sampling monitoring of hydrophobic chemicals in the salmon farm industry. Mar. Manag. 13, 585–600. Pollut. Bull. 88, 174–179. Kuivila, K.M., Hladik, M.L., Ingersoll, C.G., Kemble, N.E., Moran, P.W., Calhoun, D.L., Van Metre, P.C., Alvarez, D.A., Mahler, B.J., Nowell, L., Sandstrom, M., Moran, P., 2017. et al., 2012. Occurrence and potential sources of pyrethroid insecticides in stream Complex mixtures of pesticides in Midwest US streams indicated by POCIS time-in- sediments from seven U.S. metropolitan areas. Environ. Sci. Technol. 46, 4297–4303. tegrating samplers. Environ. Pollut. 220, 431–440. Lao, W.J., Arye, G., Ernst, F., Xu, Y.P., Bondarenko, S., Haver, D., et al., 2008. Reduction Wang, J., Taylor, A., Schlenk, D., Gan, J., 2018. Application and validation of isotope of Pyrethroid Runoff from a Commercial Nursery. Synthetic Pyrethroids. 991. dilution method (IDM) for predicting bioavailability of hydrophobic organic con- American Chemical Society, pp. 428–446. taminants in soil. Environ. Pollut. 236, 871–877. Lao, W.J., Tsukada, D., Greenstein, D.J., Bay, S.M., Maruya, K.A., 2010. Analysis, oc- Wei, Y.L., Li, H.Z., Zhang, J.J., Xiong, J.J., Yi, X.Y., You, J., 2017. Legacy and current-use currence, and toxic potential of pyrethroids, and fipronil in sediments from an urban insecticides in agricultural sediments from South China: impact of application pattern estuary. Environ. Toxicol. Chem. 29, 843–851. on occurrence and risk. J. Agric. Food Chem. 65, 4247–4254. Lao, W.J., Maruya, K.A., Tsukada, D., 2012. A two-component mass balance model for Weston, D.P., Holmes, R.W., You, J., Lydy, M.J., 2005. Aquatic toxicity due to residential calibration of solid-phase microextraction fibers for pyrethroids in seawater. Anal. use of pyrethroid insecticides. Environ. Sci. Technol. 39, 9778–9784. Chem. 84, 9362–9369. Weston, D.P., Ding, Y.P., Zhang, M.H., Lydy, M.J., 2013. Identifying the cause of sediment Laskowski, D.A., 2002. Physical and chemical properties of pyrethroids. Rev. Environ. toxicity in agricultural sediments: the role of pyrethroids and nine seldom-measured Contam. Toxicol. 174, 49–170. hydrophobic pesticides. Chemosphere 90, 958–964.

Lee, S., Gan, J., Liu, W.P., Anderson, M.A., 2003. Evaluation of Kd underestimation using Weston, D.P., Schlenk, D., Riar, N., Lydy, M.J., Brooks, M.L., 2015. Effects of pyrethroid solid phase microextraction. Environ. Sci. Technol. 37, 5597–5602. insecticides in urban runoff on chinook salmon, steelhead trout, and their in- Li, H.P., Lin, C.H., Jen, J.F., 2009. Analysis of aqueous pyrethroid residuals by one-step vertebrate prey. Environ. Toxicol. Chem. 34, 649–657. microwave-assisted headspace solid-phase microextraction and gas chromatography Witt, G., Lang, S.C., Ullmann, D., Schaffrath, G., Schulz-Bull, D., Mayer, P., 2013. Passive with electron capture detection. Talanta 79, 466–471. equilibrium sampler for in situ measurements of freely dissolved concentrations of Li, H.Z., Sun, B.Q., Chen, X., Lydy, M.J., You, J., 2013. Addition of contaminant bioa- hydrophobic organic chemicals in sediments. Environ. Sci. Technol. 47, 7830–7839. vailability and species susceptibility to a sediment toxicity assessment: application in Wolfram, J., Stehle, S., Bub, S., Petschick, L.L., Schulz, R., 2018. Meta-analysis of in- an urban stream in China. Environ. Pollut. 178, 135–141. secticides in United States surface waters: status and future implications. Environ. Sci. Li, H.Z., Cheng, F., Wei, Y.L., Lydy, M.J., You, J., 2017. Global occurrence of pyrethroid Technol. 52, 14452–14460. insecticides in sediment and the associated toxicological effects on benthic in- Xu, Y.P., Spurlock, F., Wang, Z.J., Gan, J., 2007. Comparison of five methods for mea- vertebrates: an overview. J. Hazard. Mater. 324, 258–271. suring sediment toxicity of hydrophobic contaminants. Environ. Sci. Technol. 41, Li, J.Y., Shi, W.X., Li, Z.H., Chen, Y.Q., Shao, L., Jin, L., 2018a. Equilibrium sampling 8394–8399. informs tissue residue and sediment remediation for pyrethroid insecticides in mar- Xu, Y.P., Gan, J., Wang, Z.J., Spurlock, F., 2008. Effect of aging on desorption kinetics of iculture: a laboratory demonstration. Sci. Total Environ. 616, 639–646. sediment-associated pyrethroids. Environ. Toxicol. Chem. 27, 1293–1301. Li, H.Z., Zhang, J., You, J., 2018b. Diagnosis of complex mixture toxicity in sediments: Xu, C.Y., Wang, J., Richards, J., Xu, T.B., Liu, W.P., Gan, J., 2018. Development of film- application of toxicity identification evaluation (TIE) and effect-directed analysis based passive samplers for in situ monitoring of trace levels of pyrethroids in sedi- (EDA). Environ. Pollut. 237, 944–954. ment. Environ. Pollut. 242, 1684–1692. Li, H.Z., Yi, X.Y., Chen, F., Tong, Y.J., Mehler, W.T., You, J., 2019. Identifying organic Xue, J.Y., Liao, C.Y., Wang, J., Cryder, Z., Xu, T.B., Liu, F.M., et al., 2017. Development of toxicants in sediment using effect-directed analysis: a combination of bioaccessibility- passive samplers for in situ measurement of pyrethroid insecticides in surface water. based extraction and high-throughput midge toxicity testing. Environ. Sci. Technol. Environ. Pollut. 224, 516–523. 53 (2), 996–1003. Yang, W.C., Gan, J.Y., Hunter, W., Spurlock, F., 2006a. Effect of suspended solids on Liao, C.Y., Richards, J., Taylor, A.R., Gan, J., 2017. Development of polyurethane-based bioavailability of pyrethroid insecticides. Environ. Toxicol. Chem. 25, 1585–1591. passive samplers for ambient monitoring of urban-use insecticides in water. Environ. Yang, W.C., Spurlock, F., Liu, W.P., Gan, J.Y., 2006b. Effects of dissolved organic matter Pollut. 231, 1412–1420. on permethrin bioavailability to Daphnia species. J. Agric. Food Chem. 54, Liu, W.P., Gan, J.J., Lee, S., Kabashima, J.N., 2004. Phase distribution of synthetic pyr- 3967–3972. ethroids in runoff and stream water. Environ. Toxicol. Chem. 23, 7–11. Yang, W.C., Spurlock, F., Liu, W.P., Gan, J.Y., 2006c. Inhibition of aquatic toxicity of Lydy, M.J., Landrum, P.F., Oen, A.M.P., Allinson, M., Smedes, F., Harwood, A.D., et al., pyrethroid insecticides by suspended sediment. Environ. Toxicol. Chem. 25, 2014. Passive sampling methods for contaminated sediments: state of the science for 1913–1919. organic contaminants. Integr. Environ. Assess. Manag. 10, 167–178. Yang, W.C., Hunter, W., Spurlock, F., Gan, J., 2007. Bioavailability of permethrin and Mayer, P., Parkerton, T.F., Adams, R.G., Cargill, J.G., Gan, J., Gouin, T., et al., 2014. cyfluthrin in surface waters with low levels of dissolved organic matter. J. Environ. Passive sampling methods for contaminated sediments: scientific rationale supporting Qual. 36, 1678–1685. use of freely dissolved concentrations. Integr. Environ. Assess. Manag. 10, 197–209. Yang, Y., Hunter, W., Tao, S., Gan, J., 2008. Relationships between desorption intervals Meyer, B.N., Lam, C., Moore, S., Jones, R.L., 2013. Laboratory degradation rates of 11 and availability of sediment-associated hydrophobic contaminants. Environ. Sci. pyrethroids under aerobic and anaerobic conditions. J. Agric. Food Chem. 61, Technol. 42, 8446–8451.

206 Z. Lu, et al. Environment International 129 (2019) 194–207

Yang, Y., Hunter, W., Tao, S., Gan, J., 2009. Effects of black carbon on pyrethroid You, J., Pehkonen, S., Weston, D.P., Lydy, M.J., 2008. Chemical availability and sediment availability in sediment. J. Agric. Food Chem. 57, 232–238. toxicity of pyrethroid insecticides to Hyalella azteca: application to field sediment Yi, X.Y., Li, H.Z., Ma, P., You, J., 2015. Identifying the causes of sediment-associated with unexpectedly low toxicity. Environ. Toxicol. Chem. 27, 2124–2130. toxicity in urban waterways in South China: incorporating bioavailabillity-based You, J., Brennan, A., Lydy, M.J., 2009. Bioavailability and biotransformation of sediment- measurements into whole-sediment toxicity identification evaluation. Environ. associated pyrethroid insecticides in Lumbriculus variegatus. Chemosphere 75, Toxicol. Chem. 34, 1744–1750. 1477–1482. You, J., Li, H.Z., 2017. Improving the accuracy of effect-directed analysis: the role of You, J., Harwood, A.D., Li, H.Z., Lydy, M.J., 2011. Chemical techniques for assessing bioavailability. Environ. Sci.: Processes Impacts 19 (12), 1484–1498. bioavailability of sediment-associated contaminants: SPME versus Tenax extraction. You, J., Lydy, M.J., 2006. Determination of pyrethroid, organophosphate and organo- J. Environ. Monit. 13, 792–800. chlorine pesticides in water by headspace solid-phase microextraction. Int. J. Zhang, Z., Troldborg, M., Yates, K., Osprey, M., Kerr, C., Hallett, P.D., Baggaley, N., Environ. Anal. Chem. 86, 381–389. Rhind, S.M., Dawson, J.J.C., Hough, R.L., 2016. Evaluation of spot and passive You, J., Landrum, P.F., Lydy, M.J., 2006. Comparison of chemical approaches for asses- sampling for monitoring, flux estimation and risk assessment of pesticides within the sing bioavailability of sediment-associated contaminants. Environ. Sci. Technol. 40, constraints of a typical regulatory monitoring scheme. Sci. Total Environ. 569-570, 6348–6353. 1369–1379. You, J., Pehkonen, S., Landrum, P.F., Lydy, M.J., 2007. Desorption of hydrophobic Zhang, J., You, J., Li, H.Z., Mehler, W.T., Zeng, E.Y., 2018. Particle-scale understanding of compounds from laboratory-spiked sediments measured by Tenax absorbent and cypermethrin in sediment: desorption, bioavailability, and bioaccumulation in matrix solid-phase microextraction. Environ. Sci. Technol. 41, 5672–5678. benthic invertebrate Lumbriculus variegatus. Sci. Total Environ. 642, 638–645.

207