Pollution-Induced Community Tolerance To Diagnose Hazardous

Jul 21, 2015 - ... not for PNA and TBT, confirming site-specific effects on local periphyton for ... Environmental Science & Technology 2017 51 (15), ...
0 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/est

Pollution-Induced Community Tolerance To Diagnose Hazardous Chemicals in Multiple Contaminated Aquatic Systems Stefanie Rotter,*,† Roman Gunold,‡ Sibylle Mothes,§ Albrecht Paschke,‡ Werner Brack,∥ Rolf Altenburger,† and Mechthild Schmitt-Jansen† †

Department Bioanalytical Ecotoxicology, ‡Department of Ecological Chemistry, §Department of Analytical Chemistry, ∥Department of Effect-Directed Analysis, Helmholtz-Centre for Environmental Research − UFZ, Permoserstrasse 15, 04318 Leipzig, Germany S Supporting Information *

ABSTRACT: Aquatic ecosystems are often contaminated with large numbers of chemicals, which cannot be sufficiently addressed by chemical target analyses. Effect-directed analysis (EDA) enables the identification of toxicants in complex contaminated environmental samples. This study suggests pollution-induced community tolerance (PICT) as a confirmation tool for EDA to identify contaminants which actually impact on local communities. The effects of three phytotoxic compounds local periphyton communities, cultivated at a reference (R-site) and a polluted site (P-site), were assessed to confirm the findings of a former EDA study on sediments. The sensitivities of R- and P-communities to prometryn, tributyltin (TBT) and N-phenyl-2-naphthylamine (PNA) were quantified in short-term toxicity tests and exposure concentrations were determined. Prometryn and PNA concentrations were significantly higher at the P-site, whereas TBT concentrations were in the same range at both sites. Periphyton communities differed in biomass, but algal class composition and diatom diversity were similar. Community tolerance of P-communities was significantly enhanced for prometryn, but not for PNA and TBT, confirming site-specific effects on local periphyton for prometryn only. Thus, PICT enables in situ effect confirmation of phytotoxic compounds at the community level and seems to be suitable to support confirmation and enhance ecological realism of EDA.

1. INTRODUCTION Awareness is increasing that in industrial and agricultural regions water resources are exposed to many chemicals simultaneously, which may pose a risk to aquatic ecosystems and human health.1 Addressing this complex contamination by environmental risk assessment based on chemical target analysis alone, results in a significant risk to ignore unknown and unexpected drivers of toxicity. Thus, effect-based tools are required to detect potential adverse effects together with nontarget chemical analytical tools to identify possible causes. Effect-directed analysis is a promising tool to identify drivers of toxicity in complex environmental mixtures for site-specific risk assessment. This approach combines fractionation procedures to reduce the chemical complexity of environmental samples with biotesting and finally chemical analysis for the identification of hazardous compounds.2 To validate the outcome of EDA, the hazard resulting from the identified compounds needs to be confirmed under realistic exposure conditions for higher biological levels. The EDA approach typically does not consider bioavailability, chronic exposure scenarios, sensitivity differences between species and species interactions within communities. Thus, confirmation of potential effects under realistic exposure conditions and for higher biological levels (populations or communities) is © 2015 American Chemical Society

recommended to identify the compounds actually exerting risks to the ecosystem.3 This is in agreement with the philosophy of toxicity confirmation in the closely related toxicity identification evaluation concept suggested by the U.S. Environmental Protection Agency.4,5 The quantification of pollution-induced community tolerance (PICT) is an approach that can be applied in site-specific risk assessment to detect effects on local communities. PICT is based on ecological succession processes triggered by the chronic exposure to toxicants.6 The replacement of sensitive species and individuals by more tolerant ones within a community (selection phase) results in a measurable increase of community tolerance compared to uncontaminated reference communities, representing the baseline tolerance. This increase in tolerance is quantified during a second shortterm exposure of the pre-exposed communities to the toxicant suspected to exert the selection pressure (detection phase). Therefore, PICT causally links toxicant exposure to integrated community responses, which has been shown in several field Received: Revised: Accepted: Published: 10048

March 13, 2015 July 17, 2015 July 21, 2015 July 21, 2015 DOI: 10.1021/acs.est.5b01297 Environ. Sci. Technol. 2015, 49, 10048−10056

Article

Environmental Science & Technology studies, for example, for metals,7 irgarol8 and tributyltin.9 The combination of nontarget analytical tools with a community approach enabling in situ cause-effect relationships promises to be a powerful approach for site-specific risk assessment. PICT has already been proposed as a tool to confirm EDA results.3 However, for a reliable hazard confirmation based on PICT, some uncertainties regarding the validity of PICT need to be resolved, in particular the following aspects have to be taken into account: First, the detection method and the exposure time applied during the short-term toxicity test need to be appropriate for the mode of action of the toxicant, as a long metabolic cascade between the primary target of a compound and the measured end point may limit the ability to detect PICT.10 Therefore, it is important to study the time dependence of the effects, in order to account for variations of toxicokinetic and toxicodynamic processes of toxicants. Second, co-tolerance to chemicals with similar mode of action may occur as a byproduct of tolerance to the selecting compound,6 which has to be considered for establishing causality. Finally, in complex contaminated sites several compounds might exert selection pressures leading to multiple tolerances toward different modes of action. This study aims to demonstrate the potential of the PICT approach as a confirmation tool within the framework of EDA and environmental risk assessment (ERA), identify toxicants which actually show impact on local communities and thus support site-specific risk assessment of multiple contaminated sites. To this end, a proof of principle study was performed based on a previous EDA study conducted on sediment extracts from a highly contaminated, industrial area in Bitterfeld (Germany).11 Since this study has been performed more than a decade ago and the contamination originates from industrial production that has been terminated even before, it was particularly interesting to see whether this historical burden still impacts on local periphyton communities. The authors of the study identified three compounds dominating phytotoxicity of the sediment extracts: N-phenyl-2-naphthylamine (PNA), tributyltin (TBT) and prometryn.11 Assuming partitioning of these sediment contaminants into the water phase they might still cause effects on local periphyton communities even without direct contact to sediments. Communities directly growing on the sediment might experience significantly higher exposure. However, they cannot be harvested easily and thus were not considered in our study. Prometryn, PNA and TBT have different fields of application and differ in their anticipated modes of action (properties of compounds are shown in the Supporting Information (SI), Table S1). PNA affects organisms due to reactive toxicity.12 Prometryn is a phytotoxic compound, interfering with the D1 protein of the photosystem II (PSII).13 By contrast, TBT primarily affects the energy metabolism of chloroplasts and mitochondria, interacting with proteins and membranes,14 but also inhibits the proton translocation at the Cytochrome b6fcomplex within the photosynthetic electron transport at higher concentrations (10−100 μM).15 Thus, the possibility of a cotolerance between prometryn and TBT must be taken into account in this study. While effects of prometryn have been confirmed on local communities of the investigated site,16,17 possible effects of PNA and TBT on local communities have not been analyzed, up to now. Moreover, the toxicants have been identified in an exhaustive sediment extract, which does not reflect bioavailable concentrations in the water phase. Thus, effects need to be confirmed

under realistic exposure conditions taking into account the different modes of action and the time dependency of effects on local algal communities. Therefore, different measures of photosynthesis were used to enable an appropriate effect evaluation of the selected compounds.

2. MATERIALS AND METHODS 2.1. Sampling Sites and Periphyton Colonization. This study was carried out in two streams of the river Elbe basin in the area of Bitterfeld (Germany) from May to June 2011. With regard to the findings of Brack et al.11 the creek Spittelwasser was chosen as polluted sampling site (in the following called Psite). The reference site (R-site) was located at the river Mulde, 9 km upstream of the junction with the contaminated creek. Periphyton was colonized on circular glass discs (1.77 cm2) mounted in plastic racks according to Blanck.18 Seven racks, each supporting 100 glass discs were exposed parallel to the current, 10−15 cm below the water surface at the R- and the Psite, respectively. After 7, 14, 21, and 28 days, 20 glass discs were collected for characterization of periphyton communities (biomass, algal class and diatom composition). After 20, 21, 22, and 28 days about 120 glass discs were retrieved from both sites for short-term toxicity tests. For acclimatization, periphyton was incubated in the laboratory at 20 °C and a photosynthetic photon flux density (PPFD) of about 130 μmol photons m−2 s−1 for 24 h before testing. 2.2. Physico-Chemical Water Analyses. Physico-chemical water parameters (conductivity, light attenuation, oxygen concentration, pH, and current velocity) were measured directly on site. For further chemical analyses, water samples were taken, kept cool during transport to the laboratory and stored at 4 °C until analysis (maximum 24 h). Quantification of dissolved ions (Cl−, F−, NO2−, NO3−, SO42−: ion chromatography; NH4+, PO43−: photometry; Ca2+, K+, Mg2+, Na+, Si4+: ICP-OES) and trace elements (As3−, Cd2+, Cu2+, Fe2+, Mn2+, Zn2+: ICP-OES) was conducted following standard techniques. Furthermore, subsamples were used for quantification of prometryn, PNA and TBT. 2.2.1. Prometryn and PNA Analyses. A solid phase extraction (SPE) was performed to analyze prometryn and PNA concentrations, using two replicates per sampling date and site. The samples were processed and analyzed in accordance to Rotter et al.17 using GC-MS for quantification. Concentrations were corrected by using deuterated benzo[a]pyrene-d12 as injection standard. 2.2.2. TBT Analyses. For quantification of TBT, water samples were filtered through glass-fiber filters (Whatman, GF/ F) and extracted using solid-phase microextraction (SPME) within 48 h after sampling. For processing of the samples, 9.5 mL acetate buffer (pH 4.75) and 20 μL sodium propyl borate (NaBPr4; CAS RN: 45067−99−0; Merseburger Spezialchemikalien, Schkopau, Germany) solved in tetrahydrofuran (2%) (CAS RN: 109−99−9; Merck, Darmstadt, Germany) were added to 1 mL of the water sample. Afterward, the solution was stirred for 10 min. The propylated butyltin compounds were extracted by SPME using a 100 μm polydimethylsiloxane (PDMS)-coated fused silica fiber (Supelco, Bellefonte, PA) at 40 °C for 30 min. Subsequently, the fiber was automatically injected in a 6890 gas chromatograph (Hewlett-Packard, Germany) with a HP-5MS capillary column (30 m × 250 μm × 0.25 μm, Agilent, Santa Clara, CA) and thermally desorbed at 250 °C under constant helium flow (2.5 mL min−1). The separated compounds were detected with an atomic emission 10049

DOI: 10.1021/acs.est.5b01297 Environ. Sci. Technol. 2015, 49, 10048−10056

Article

Environmental Science & Technology

TBT were conducted with uncontaminated filtered freshwater at three consecutive days after 3 and 4 weeks of colonization, respectively. Due to technical problems [14C]-incorporation tests were not measured after 4 weeks. The tests for each compound were conducted within 10 h. 2.4.1. Pulse-Amplitude Modulation Fluorometry, Measurements. PAM fluorescence measurements are based on variable chlorophyll a fluorescence as a measure of the efficiency of the electron transport of the PSII. Periphyton of each site was exposed to six different concentrations of prometryn (1.26 × 10−4 to 126 μmol L−1, CAS RN: 7287−19−6, Riedel-de Haën, Seelze, Germany), tributyltin chloride (1.11 × 10−4 to 11 μmol L−1, CAS RN: 1461−22−9, Merck, Darmstadt, Germany) and PNA (5.66 × 10−2 to 28 μmol L−1, CAS RN: 135−88−6, Sigma-Aldrich, Germany), respectively. Stock solutions of compounds were prepared in dimethyl sulfoxide (DMSO; CAS RN: 67−68−5, Merck, Darmstadt, Germany). Each test consisted of six concentrations, a solvent control containing DMSO and an untreated control with three colonized replicate glass discs, respectively. The DMSO concentration was kept at 0.1% and did not show effects in the solvent controls. Glass discs colonized with periphyton were incubated in small glass vials on a rotary shaker and under a PPFD of 110 μmol m−2 s−1 at 20 °C. Preliminary tests were conducted to decide for appropriate incubations times for each compound, respectively. The maximum and effective photosynthetic quantum yield were determined by applying the saturation pulse method. To calculate the maximum quantum yield (ΦPSII), the minimal (F0) and maximum level of fluorescence (Fm) were measured using dark adapted periphyton (eq 1). The effective quantum yield (Φ′PSII) was estimated under actinic light, driving the photosynthetic electron transport, by measuring the steady state fluorescence (F0′) and the maximum fluorescence in the light (Fm′; eq 2).

detector (G2350A, Joint Analytical Systems, Moers, Germany) under optimized conditions. 2.3. Passive Sampling. The additional use of passive sampling techniques enables a continuous monitoring of bioavailable compounds and represents the exposure history of periphyton during the growth period. Two different passive sampling configurations were used in order to detect timeweighted average (TWA) concentrations of prometryn, PNA and TBT. The samplers were fixed in stainless steel cages and were exposed next to the biofilm racks. 2.3.1. Polar Organic Chemical Integrative Sampler (POCIS). POCIS were used for monitoring of PNA and prometryn. Sampler design and operating principles have been described by Alvarez et al.19 The composition of POCIS as well as the extraction, calibration, quantification of the compounds and the calculation of TWA concentrations were conducted as described previously.17,20 Briefly, calibration was conducted at 20 °C and a water flow of 0.15 m s−1, resulting in sampling rates of 0.59 L d−1 for prometryn and 0.57 L d−1 for PNA. A total of four samplers were exposed at each sampling site for 21 days. After sampling, POCIS were covered with aluminum foil, kept cool, and were stored at −20 °C until analysis. 2.3.2. Chemcatcher. For analysis of TBT, the Chemcatcher was used according to the description of Aguilar-Martı ́nez et al.21 C18 Empore Extraction discs (47 mm diameter; 3M, St. Louis) were used as receiving phase, covered with a diffusionlimiting cellulose acetate membrane (0.45 μm pore size; Sartorius Stedim Biotech GmbH, Göttingen, Germany). The prepared samplers were stored at 4 °C until exposure at the sampling site (max. 92 h). At each sampling site, three Chemcatcher bodies were exposed in two periods of 14 days each. Directly after retrieval, each sampler was filled with water from the sampling site and stored at 4 °C until extraction. For the extraction, solvents (methanol, hexane) and reagents (acetic acid, sodium acetate, tripropyltin) were of analytical grade and were purchased from Merck (Darmstadt, Germany). The derivatization agent sodium tetraethylborate (NaBEt4, CAS RN: 15523−24−7) was obtained from Sigma-Aldrich (Steinheim, Germany) with a purity of at least 97%. For quantification of TBT, a GC-AED device was used as described for TBT analyses of water samples (see 2.2.2 TBT analysis). However, here aliquots of 1 μL per sample were injected splitless at a temperature of 250 °C. The column temperature was kept at 60 °C for 1 min, increased with a rate of 20 °C min−1 up to 180 °C, then with 30 °C min−1 to 300 °C, where the temperature was kept constant for 3 min. All TBT concentrations were corrected using tripropyltin as internal standard. The TWA concentrations were calculated based on the sampling rate (0.056 L d−1) obtained at 18 °C and low turbulence by Aguilar-Martı ́nez et al.21 2.4. Analyses of Chronic Effects on Periphyton. In order to assess the chronic effects of the selected toxicants, periphyton communities were characterized (algal class composition, biomass and diatom composition) weekly and community tolerance was quantified in short-term inhibition tests for the detection of PICT. As the studied compounds are phytotoxic, the effect assessment focused on the autotrophic community, potential impacts on heterotrophs and interactions between algae and bacteria were not in the scope of the study. Preliminary tests were conducted to determine appropriate exposure durations and concentration ranges for each compound. Short-term toxicity tests of prometryn, PNA and

⎛ F ′m − F ′0 ⎞ ⎛ F − F0 ⎞ ⎟⎟ ϕPSII = ⎜ m ⎟⎟ and ϕ′PSII = ⎜ ⎝ F ′m ⎠ ⎝ Fm ⎠

(1,2)

Samples were measured in time series with four observations respectively, using the MAXI-Imaging-PAM (Walz, Effeltrich, Germany). The average of these measurements was used for calculation of photosynthesis inhibition. 2.4.2. [14C]-Incorporation. The incorporation of radiolabeled sodium bicarbonate into macromolecules was applied as a more integrated measurement of photosynthetic rate. Sample preparation, concentration ranges and total incubation times were the same as for PAM fluorometry measurements (see Section 2.4.1). Due to a higher biological variation of the [14C]-incorporation, five colonized replicate glass discs per concentration and control were used. In addition, one set of samples remained unlabeled and the rate of dark fixation was estimated by keeping further controls in the dark. Furthermore, five samples were inactivated with formol (0.25% final concentration) to detect the abiotic CO2 fixation of the periphyton. After preincubation for 30 min to TBT and prometryn and 23 h 30 min to PNA, respectively, 25 μL Na[H14]CO3 (PerkinElmer LAS, Rodgau, Germany) was added to the glass vials giving a final activity of 0.0625 μCi ml−1 (2.31 kBq ml−1). Carbon fixation was terminated after 30 min by adding 50 μL formol (10%) to each sample. The samples were processed according to Schmitt-Jansen and Altenburger.22 The number of [14C]-disintegrations per minute were measured in a 10050

DOI: 10.1021/acs.est.5b01297 Environ. Sci. Technol. 2015, 49, 10048−10056

Article

Environmental Science & Technology

Prometryn and PNA showed higher water concentrations at the polluted site than at the reference site throughout the duration of the experiment (Figure 1). Taking into

Wallac 1414 liquid scintillation counter (PerkinElmer Wallac GmbH, Freiburg, Germany). 2.4.3. Calculation of Effect Parameters. For each shortterm toxicity test mean values of controls were set to 100% of activity and relative inhibition of each exposure concentration was calculated to model concentration−response relationships, using log−logistic analysis. Maximum response (Amax) was set to 100%, minimal response (Amin) to zero, y represents the effect, x the concentration, x0 the concentration at median efficacy (EC50) and p the slope of the curve (eq 3). γ = A max +

A min + A max 1 + (x /x0) p

(3) 14

However, there are exceptions for some [ C]-incorporation tests, showing considerable (−25%) relative effects below 0%. In this case, the mean of the relative effect values of lowest tested concentration was used as lower asymptote and EC50 was recalculated by using parameters of the respective concentration−response curve. Analyses were performed using nonlinear regression least-squares curve fitting with the software OriginPro 8G (Microcol Software Inc., Northhampton, MA). Statistical differences between EC50s of Rand P-communities were tested using an unpaired t test. 2.4.4. Algal Class Composition. Algal class composition was analyzed based on fluorescence measurements using a fourwavelength-excitation PHYTO−PAM fluorometer (Heinz Walz GmbH, Effeltrich, Germany). The signature pigments of diatoms, chlorophytes and cyanobacteria allow for the discrimination by multiwavelength excitation. Fluorescence was measured for three spots per glass disc with five replicates per site and sampling date according to Schmitt-Jansen and Altenburger.22 2.4.5. Biomass. Chlorophyll a concentrations were used as a measure for autotrophic biomass. The extraction and analysis of lipophilic pigments were conducted according to SchmittJansen and Altenburger.23 Five colonized glass discs per site and sampling date were analyzed using reverse-phase high performance liquid chromatography (HPLC). Chlorophyll a was calibrated using standards supplied by Sigma-Aldrich (Steinheim, Germany). 2.4.6. Taxonomic Analyses of Diatoms. After 3 and 4 weeks of colonization, ten glass discs per sampling site were preserved in 5% formaldehyde for identification of diatom species. Preparation and identification of the samples was conducted according to Krammer and Lange-Bertalot.24 At least 530 valves per sample were counted and identified using light microscopy. For each sample the Shannon Diversity Index, Evenness, Specific pollution sensitivity index (IPS), Trophic index and Saprobic index were calculated using Omnidia software v8.1.25 Furthermore, the Index of Salinity was calculated according to Ziemann et al.26

Figure 1. Concentrations of prometryn, TBT and PNA in nmol L−1 detected in weekly spot sampling (two replicates) as well as from passive sampling with POCIS (four replicates) and Chemcatcher (2 × 3 replicates) at the reference (R-site; circles) and the polluted (P-site; triangles) site. Solid lines represent time-weighted average concentrations for the P-site, dotted lines for the R-site.

consideration all spot samplings, Prometryn concentrations detected at the P-site had a mean of 1.67 nmol L−1 and were significantly (t test, p < 0.05) higher than at the R-site (0.006 nmol L−1). In contrast, mean concentrations of TBT with 0.31 nmol L−1 at the P-site and 0.24 nmol L−1 at the R-site were similar. The average PNA concentration at the P-site (0.026 nmol L−1) was significantly higher than at the R-site (0.007 nmol L−1; t test: p < 0.05). TWA concentrations derived from passive samplers confirmed the concentration trend of spot sampling (Prometryn > TBT > PNA concentration) for both sites. However, TWA concentrations were two to three times lower than mean concentrations detected in spot samples, unlike spot sampling, particle-/DOC bound fractions are not taken up by passive samplers. TWA concentrations from POCIS (4 replicates) and Chemcatcher (2 × 3 replicates) varied strongly for the R-site, resulting in higher standard deviations (SI Table S.2). Based on the Grubbs outlier test one Chemcatcher replicate was excluded from data evaluation. Nevertheless, TBT concentrations were below 0.01 nmol L−1 in four out of five remaining Chemcatcher replicates, whereas the fifth extract showed a concentration of 0.56 nmol L−1 for the Rsite. 3.2. Structural Characterization of Periphyton Communities. Periphyton communities of the R- and P-site

3. RESULTS 3.1. Physico-Chemical Characterization of the Sampling Sites. On average, conductivity at the P-site (1934 μS cm−1) was four times higher than at the R-site (560 μS cm−1), mainly due to higher concentrations of Cl−, SO42−, Ca2+, K+ and Na+ (SI, Table S.2). Flow velocity, pH, light attenuation as well as the concentration of O2 and F− were similar at both sites, whereas temperature, NO3− and PO43− were higher at the R-site. The concentrations of As3−, Cd2+, Cu2+, and Zn2+ were below detection limits at both sites. 10051

DOI: 10.1021/acs.est.5b01297 Environ. Sci. Technol. 2015, 49, 10048−10056

Article

Environmental Science & Technology Table 1. Community Parameters of Periphyton from R- and P-Site after 3 and 4 Weeks of Colonizationa R-site biological parameters

third week

−2

4.08 ± 1.49 0.2 ± 0.4 36.2 ± 2.8 63.7 ± 2.8 49 2.24 0.40 15.2 2.5 1.9 1.3

chlorophyll a [μg cm ] cyanobacteria [%] chlorophytes [%] diatoms [%] taxa richness [n] Shannon diversity evenness IPS trophic index saprobic index salinity index

*

P-site fourth week

third week

± 2.40 ± 6.6 ± 17.8 ± 11.7 54 2.20 0.38 15.1 2.5 1.9 1.6

2.29 ± 0.46 0.5 ± 0.7 40.2 ± 4.1 59.3 ± 3.5 79 2.23 0.35 17.4 1.8 1.8 1.0

4.87 6.1 38.7 55.2

fourth week *

2.27 ± 0.40 7.5 ± 6.7 34.1 ± 11.5 55.5 ± 7.5 70 2.30 0.38 17.3 1.9 1.9 1.4

Chlorophyll a concentrations and relative proportions of algal classes are mean values derived from five replicate glass discs (±represent standard deviations). The taxa richness and species indices for Shannon diversity index, evenness, specific pollution sensitivity index (IPS), trophic index, saprobic index and the salinity index are based on diatom abundances. (* indicate significant differences between the sampling sites). a

Table 2. Effective Concentrations (EC50) in μmol L−1 of Prometryn, TBT and PNA for Periphyton Communities Colonized at the Reference (R-site) and the Polluted Site (P-site), Determined with Three Measures of Photosynthesis after 3 and 4 Weeks of Colonizationa prometryn −1

EC50 [μmol L ] third week ΦPSII PAM Φ′PSII PAM [14C]-incorporation

fourth week ΦPSII PAM Φ′PSII PAM

R-site 1.797 ± 0.489 0.035 ± 0.004 0.054 ± 0.004

4.146 ± 1.113 0.036 ± 0.001

* * *

* *

TBT P-site

R-site

6.896 ± 2.366 0.072 ± 0.011 0.096 ± 0.01

0.323 ± 0.028 0.139 ± 0.009 0.011 ± 0.003

22.218 ± 5.606 0.129 ± 0.011

0.445 ± 0.055 0.083 ± 0.016

* * *

PNA P-site

R-site

P-site

0.148 ± 0.022 0.006 ± 0.002 0.101 ± 0.014

(198.72) (±78.46) (173.35) (±104.77) 6.034 ± 3.926

(175.90) (±93.00) (69.84) (±69.84) 4.197 ± 0.658

0.336 ± 0.033 0.084 ± 0.014

(167.96) (±43.39) (31.184) (±7.786)

Median efficiencies were modeled from concentration−response relationships using log−logistic data analyses, ± represent their standard errors. EC50 values in brackets represent approximate values, since maximum inhibition was less than 50% in the concentration−response curve. (ΦPSII = maximum quantum yield; Φ′PSII = effective quantum yield; * indicate significant differences between EC50s of R- and P-site). a

3.3. Tolerance Measurements of Periphyton Communities. The preliminary short-term toxicity tests revealed time-dependent effects on periphyton (data and detailed description are shown in the SI, Table S4.). Therefore, exposure durations of 24 h were chosen for PNA to derive effects reliable for concentration−response modeling, while prometryn and TBT showed best results after 1 h. The sensitivities of R- and P-communities to prometryn, PNA and TBT were detected with three different measures of photosynthesis. Median efficacies (EC50s) are shown in Table 2 and model parameters in Table S5. of the SI. The concentration−response relationships based on the effective quantum yield measurements after 4 weeks are illustrated in SI Figure S1. The EC50 values for prometryn, calculated from the different measures of photosynthesis, varied, showing the lowest values for the effective quantum yield, closely followed by [14C]incorporation, whereas the maximum quantum yield gave 50− 95 times higher EC50 values. Nevertheless, in all experiments periphyton from the P-site was about two (Φ′PSII and [14C]incorporation) to 3.8 (ΦPSII) times more tolerant than the R-

differed in biomass, taxa richness and IPS classification, while other biological parameters were similar (Table 1). Rcommunities had significantly higher chlorophyll a concentrations than P-communities (t test, p < 0.05), indicating a higher autotrophic biomass after 4 weeks (Table 1). The relative abundances of algal classes showed no significant differences (SI Table S.3) and were comparable, consisting of about 4% cyanobacteria, 37% chlorophytes and 58% diatoms. In total, 148 diatom species were determined, a higher taxa richness of diatoms was found for P-communities, but Shannon diversity and Evenness were in the same range. The Pcommunities were dominated by Achnanthes minutissima (74.8%) and the R-communities by Cocconeis placentula (64.4%) and A. minutissima (17.2%). IPS values of diatom communities were higher at the P-site than at the R-site. Furthermore, the R-site and the P-site differed in the Trophic index, whereas the Saprobic and the Salinity index were similar. With the exception of cyanobacteria, differences for all biological measures were low between the third and the fourth week, therefore a climax state of the periphyton communities can be assumed. 10052

DOI: 10.1021/acs.est.5b01297 Environ. Sci. Technol. 2015, 49, 10048−10056

Article

Environmental Science & Technology community. These differences in tolerance were significant (SI Table S6). Data retrieved after 4 weeks of colonization showed a 3.6−5.4 times higher tolerance of the P-site community, confirming the results for 3 weeks of colonization. Sensitivity of effective quantum yield (Φ′PSII) and [14C]incorporation to TBT were in the same range, whereas maximum quantum yield gave again highest EC50 values (two to 26 times higher). For the measures of maximum and effective quantum yield, reference communities were 2−24 times more tolerant than P-communities. However, the quantification of [14C]-incorporation revealed communities of the P-site being nine times more tolerant than R-communities. After 4 weeks of colonization, measurements of maximum and effective quantum yield showed similar tolerances of both communities. Significant differences were found for all measures after 3 weeks, but not after 4 weeks (SI Table S6). Thus, R- and P-communities showed different trends of TBT tolerance for different measures of photosynthesis and time points. However, there is no systematic effect as found for prometryn. For PNA the [14C]-incorporation was the most sensitive measure of photosynthesis inhibition. For maximum and effective quantum yield, maximum inhibition of photosynthesis did not exceed 50% and replicates showed high variations. Therefore, the calculated EC50 values represent an approximate estimation. However, the highest applied test concentration caused similar relative photosynthesis inhibition of R- and Pcommunities, indicating similar sensitivities. With regard to [14C]-incorporation, R-communities are slightly (1.4 times) more tolerant than P-communities, but were not significantly different (SI Table S6). The short-term toxicity test for the Psite after 4 weeks could not be evaluated, due to high variations of replicates and maximum inhibitions below 40%. For this reason, a comparison of data between the third and the fourth week was not possible.

For prometryn, all measures determined effects on photosynthesis within the first hour, which is consistent with the specific mode of action of prometryn. TBT alters the mitochondrial structure and function and inhibits oxidative phosphorylation as well as the synthesis of ATP.14 It only affects photosynthesis directly in concentrations exceeding exposure concentrations tested in this study.15 For this reason, it was assumed the that the more integrated measure of [14C]-incorporation is more directly affected by TBT than the effective and maximum quantum yield, but all measures of photosynthesis indicated a fast effect development. Thus, for prometryn and TBT all applied measures of photosynthesis are suitable for the detection of tolerance, but effective quantum yield was more sensitive than maximum quantum yield. PNA affects organisms due to a reactive toxicity, which leads to an irreversible and thus slowly cumulating damage.12 Consequently, photosynthesis might not be directly affected by PNA, which is also indicated by the extended exposure time required for effect assessment in our study. For PNA, [14C]-incorporation was the most sensitive measure, whereas PAM fluorescence measures were not suitable with acceptable exposure times. According to Eriksson,10 [14C]incorporation might be more qualified to detect effects of compounds with modes of action different from inhibition of photochemistry. Accordingly, end points have to be chosen according to the mode of action of the compound to obtain accurate tolerance values.6 For a reactive mode of action an integrated end point might be more reliable. In consequence, a choice of integrated and several more specific end points is required for a reliable PICT detection. 4.2. Ecological Status of Sampling Sites. Beside toxicants, the abiotic environment influences species composition and thus affects the responses of periphyton to toxic exposure.28 For instance, the higher concentration of phosphate and nitrate at the R-site might partly cause the higher biomass of communities from the R-site and the different categorization based on the trophic index for diatoms. Whereas the R-site was classified as eutrophic containing a high amount of nutrients, the P-site had a moderate nutrient level (mesotrophic). This might be due to a higher agricultural impact at the R-site. Both sites showed moderate organic pollution levels according to the Saprobic index. The differences in chloride concentrations and conductivity were not reflected in the salinity index, indicating typical freshwater conditions (β-oligohalob) at both sites. Also diversity and evenness of diatom communities were in the same range. As phytobenthos is one of the biological quality elements of the WFD, diatom indices such as IPS are frequently used indicators for the assessment of the ecological status within the WFD. Beside the eutrophic status and the moderate pollution level, the P-site was categorized as a site with high ecological status, whereas the R-site showed a good ecological status based on the IPS values and the limits proposed by Eloranta and Soininen.29 This shows the importance to combine saprobic index, trophic index, and species composition metrics for the assessment of ecological status of phytobenthos within the WFD.30 4.3. Exposure and Effects of Studied Compounds. The objective of this study was to identify effects of phytotoxic, sitespecific compounds on local communities to demonstrate the potential of PICT within ERA. Therefore, a polluted and a reference site were chemically characterized and their respective communities biologically analyzed.

4. DISCUSSION 4.1. Suitability of Chosen End Points for PICT Detection. In order to create a sound basis for the hazard evaluation of PICT responses in a field approach, the suitability of the chosen end points is of high relevance. Depending on the mode of action of the compounds measurable effects might occur within different time ranges.27 In general, one can assume that a long metabolic cascade between the measured end point in short-term toxicity testing and the primary target of the compound leads to a delayed detection of the effect.10 Furthermore, the effective concentration is highly dependent on the sensitivity of the end point. An end point which is indirectly affected by the compound might underestimate the toxicity of the compound, if the exposure duration is not sufficient for a solid effect development. On the other hand, the induction of new selection processes within the short term test, such as significant growth of tolerant species, would influence the PICT detection and must be avoided.6 On this account, the exposure time of 24 h should not be further extended in the detection phase. The effective and the maximum quantum yield indicate the efficiency of the electron transport in the light reaction, whereas [14C]-incorporation gives an estimation of the light independent (dark) reaction within the Calvin cycle. In comparison to PAM measures, [14C]-incorporation represents a more integrated measure of photosynthesis with more and closer links to other parts of the metabolism.10 10053

DOI: 10.1021/acs.est.5b01297 Environ. Sci. Technol. 2015, 49, 10048−10056

Article

Environmental Science & Technology

detected for TBT. Furthermore, the tolerance toward prometryn did not lead to co-tolerance toward TBT, confirming that different modes of action of compounds may require different tolerance mechanisms at the level of a community response. The PNA exposure at the P-site (0.026 nmol L−1) was higher than at the R-site by a factor of 3. However, concentrations at both sites were low, being 3−4 orders of magnitude lower than an EC 50 (0.153 μmol L −1 ) detected for Scenedesmus vacuolatus.12 Furthermore, a metabolomics approach studying the effect of PNA on Scenedesmus vacuolatus reported first effects on metabolite concentrations for exposures of 0.007− 0.228 μmol L−1 and growth-related effects from 0.456 μmol L−1 on.37 Thus, the chronic exposure concentration in this study is still 2−3 orders of magnitude lower than the concentrations triggering first effects on the metabolome level. In comparison to single species tests, median efficacies detected for periphyton communities were with 4−6 μmol L−1 about one magnitude higher than for single species tests. The low exposure concentration found in the field and the result that no sensitivity differences between R- and P-communities were detected, led us to the conclusion that PNA did not affect local communities in the water phase. As a result, out of three occurring phytotoxic toxicants, sitespecific effects at the community level could only be confirmed for prometryn. Since the EDA study has been performed more than a decade ago, it is expected that concentrations have declined since then. Further, the underlying EDA study used a sediment extract, derived from exhaustive extraction methods, containing total amounts of contaminants, therefore neglecting that bioaccessibility depends on desorption kinetics of compounds, which may be very slow in situ.38 Thus, compounds found in the sediment extract are not necessarily present in adverse concentrations in the water phase. In comparison to PNA and TBT, prometryn has the lowest log KOW, is readily soluble (SI Table S1) and stable in water.17 In contrast, TBT has a high lipophilicity and is accumulating in sediments, but slowly degrades to less toxic dibutyltin (DBT) and monobutyltin (MBT) in water,21 also detected in our study (data not presented). PNA is rather lipophilic with a log KOW of about 4.5 and expected to exhibit slow desorption kinetics into the water phase.39 Accordingly, the physico-chemical parameters explain the low concentrations of TBT and PNA measured in water as well as the non-detectable effects on the communities. However, for benthic algae being in direct contact with sediments, partitioning processes in the sedimentwater-biota system might lead to bioavailable equilibrium concentrations causing effects. 4.4. PICT as Diagnostic Tool for Effect Confirmation. This study shows that PICT is well-suited to support the confirmation of community impacts of phytotoxic compounds identified by EDA. Thus, PICT is a valuable complement of the array of confirmation approaches suggested for EDA3 and TIE.4 PICT can also be used to monitor recovery effects after management measures such as the ban of contaminants.40 Limitations may be seen in the current restrictions of toxicological end points and test organisms. When used as a confirmation step in sediment EDA, advancing exposure regimes from periphyton communities from exposure to water only, toward direct contact to contaminated sediment would further increase environmental realism. Co-tolerances for chemicals with similar modes of action might limit the PICT approach, but were not found in this study. Furthermore,

R- and P-communities showed a similar biodiversity and algal class composition, but differed at the species level of diatoms. A previous study revealed that the sensitivity of the whole community was positively correlated with the abundance of Achnanthes minutissima,17 which dominated the P-communities of our study. The higher biomass of the R-site, might limit the transfer of contaminants in biofilms,31 however, decreasing sensitivities with increasing biomass were shown for metals32 but not for organic toxicants so far. Prometryn concentrations were significantly higher (1.66 nmol L−1) at the P-site and were about two times lower than the reported no-observed effect concentration (NOEC) for algae of 3.4 nmol L −1 (reproduction of Scenedesmus vacuolatus).33 Despite the relatively low prometryn concentrations and the higher biomass of R-communities, periphyton communities colonizing the P-site were significantly and 2−5.4 times more tolerant than R-communities. This is in accordance with result of a previous field study at the same sampling site, reporting a five times higher tolerance as compared to communities of the reference site.16 A recovery study performed at the same sampling site, demonstrated structural and functional recovery within 24 days after transferring periphyton from the P-site to the R-site.17 Furthermore, a microcosm study conducted by Schmitt-Jansen and Altenburger23 showed an increase of tolerance by a factor of 3−6 for communities exposed to prometryn concentrations equal to or higher than 41 nmol L−1, but not for concentrations below. This study shows that a low but chronic exposure to 1.66 nmol L−1 is still affecting periphyton communities in the field and triggers community changes such as toxicant-induced succession, resulting in a PICT response. The low effect thresholds detected in the field might be due to a more restricted species input in microcosm studies and a possible underestimation of field exposure or confounding factors.34 Therefore, this in situ PICT study was more sensitive than studies conducted in microcosms and revealed effects of prometryn in a site categorized as site with a high ecological status (based on IPS). Thus, diatom indices are useful to indicate trophic and saprobic differences between sites, but do not necessarily reflect effects of chemical contaminants on periphyton. Considering this, effect-based tools such as PICT are required to complement the taxonomy-based tools currently applied in the WFD. With regard to TBT, exposure at the R- and P-site was in the same range (0.24−0.31 nmol L−1), which was about hundred times higher than the maximum allowable concentration environmental quality standard (MAC-EQS) of 0.0025 nmol L−1 defined by the WFD based on retarded larvae development in crustacean.35 Due to the similar exposure levels at both sites, no baseline tolerance of unaffected communities was detectable and differences in PICT responses could therefore not be expected. After 4 weeks of colonization, R- and P-communities indeed showed similar sensitivities. With regard to the findings of Blanck and Dahl,9 exposure concentrations in this study were in the range of the no-effect concentrations (0.3−0.5 nM) estimated for marine periphyton. Furthermore, a similar study using [14C]-incorporation for the quantification of periphyton sensitivity to TBT, detected an EC50 value of 0.012 μmol L−1 for reference communities,36 which is in accordance to our study (0.01 to 0.1 μmol L−1). As the exposure concentrations occurred below the no-effect concentration estimated for periphyton, local periphyton seems unaffected by TBT at the site of investigation. Thus, no clear PICT response has been 10054

DOI: 10.1021/acs.est.5b01297 Environ. Sci. Technol. 2015, 49, 10048−10056

Article

Environmental Science & Technology Blanck and Wängberg41 showed that co-tolerance indicates highly comparable toxicity and tolerance mechanisms and is not common. Moreover, several PICT studies were able to link exposure to chemical mixtures to community-level effects in multiple contaminated sites.7,42,43 Nevertheless, PICT studies should be underpinned with a chemical characterization of the sampling site to be able to address potential co-tolerance or mixture effects. In conclusion, PICT is a sensitive approach, which enables in situ effect confirmation of toxic compounds. Hazard confirmation is not only implemented in the tiered confirmation of an EDA study, it is also an important part of ERA in general and in particular in site-specific risk assessment. To our knowledge, there are no tools available so far, which causally link exposure of compounds in the field to effects on algal communities to allow an effect-based in situ hazard confirmation on higher biological levels. This becomes even more important with regard to increasing multiple contaminations in aquatic ecosystems worldwide. Thus, PICT might be used as a diagnostic tool for contaminated sites where suspected compounds are known already, but it is, as shown by Pesce et al.,43 also able to determine multiple tolerances from mixtures of putative toxicants.

■

Office of Research and Development, U.S. Environmental Protection Agency: Duluth, U.S.. (5) Burgess, R. M.; Ho, K. T.; Brack, W.; Lamoree, M. Effectsdirected analysis (EDA) and toxicity identification evaluation (TIE): Complementary but different approaches for diagnosing causes of environmental toxicity. Environ. Toxicol. Chem. 2013, 32 (9), 1935− 1945. (6) Blanck, H. A critical review of procedures and approaches used for assessing pollution-induced community tolerance (PICT) in biotic communities. Hum. Ecol. Risk Assess. 2002, 8 (5), 1003−1034. (7) Fechner, L. C.; Gourlay-France, C.; Bourgeault, A.; TusseauVuillemin, M. H. Diffuse urban pollution increases metal tolerance of natural heterotrophic biofilms. Environ. Pollut. 2012, 162, 311−318. (8) Blanck, H.; Eriksson, K. M.; Gronvall, F.; Dahl, B.; Guijarro, K. M.; Birgersson, G.; Kylin, H. A retrospective analysis of contamination and periphyton PICT patterns for the antifoulant irgarol 1051, around a small marina on the Swedish west coast. Mar. Pollut. Bull. 2009, 58 (2), 230−237. (9) Blanck, H.; Dahl, B. Pollution-induced community tolerance (PICT) in marine periphyton in a gradient of tri-n-butyltin (TBT) contamination. Aquat. Toxicol. 1996, 35 (1), 59−77. (10) Eriksson, K. M. Impact of antifouling compounds on photosynthesis, community tolerance and psbA genes in marine periphyton. Ph.D Dissertation, University of Gothenburg, 2008. (11) Brack, W.; Altenburger, R.; Ensenbach, U.; Moder, M.; Segner, H.; Schuurmann, G. Bioassay-directed identification of organic toxicants in river sediment in the industrial region of Bitterfeld (Germany) - A contribution to hazard assessment. Arch. Environ. Contam. Toxicol. 1999, 37 (2), 164−174. (12) Altenburger, R.; Brack, W.; Greco, W. R.; Grote, M.; Jung, K.; Ovari, A.; Riedl, J.; Schwab, K.; Kuster, E. On the mode of action of Nphenyl-2-naphthylamine in plants. Environ. Sci. Technol. 2006, 40 (19), 6163−6169. (13) Huppatz, J. L. Quantifying the Inhibitor-Target Site Interactions of Photosystem II Herbicides. Weed Sci. 1996, 44 (3), 743−748. (14) Fent, K. Ecotoxicology of organotin compounds. Crit. Rev. Toxicol. 1996, 26 (1), 3−117. (15) Klughammer, C.; Heimann, S.; Schreiber, U. Inhibition of cytochrome b563-oxidation by triorganotins in spinach chloroplasts. Photosynth. Res. 1998, 56 (2), 117−130. (16) Schmitt-Jansen, M.; Reiners, M.; Altenburger, R. Biozönotisches Testverfahren (PICT-Konzept) - Analyse von Schadstoff-induzierten Effekten in Gewässern mit autotrophen Aufwuchszönosen. Umweltwiss. Schadst.-Forsch. 2004, 16, 85−91. (17) Rotter, S.; Sans-Piche, F.; Streck, G.; Altenburger, R.; SchmittJansen, M. Active bio-monitoring of contamination in aquatic systems–an in situ translocation experiment applying the PICT concept. Aquat. Toxicol. 2011, 101 (1), 228−236. (18) Blanck, H. A Simple, community level, Ecotoxicological test system using samples of periphyton. Hydrobiologia 1985, 124 (3), 251−261. (19) Alvarez, D. A.; Petty, J. D.; Huckins, J. N.; Jones-Lepp, T. L.; Getting, D. T.; Goddard, J. P.; Manahan, S. E. Development of a passive, in situ, integrative sampler for hydrophilic organic contaminants in aquatic environments. Environ. Toxicol. Chem. 2004, 23 (7), 1640−1648. (20) Mazzella, N.; Debenest, T.; Delams, F. Comparison between the polar organic integrative sampler and the solid-phase extraction for estimating herbicide timeweighted average concentrations during a microcosm experiment. Chemosphere 2008, 73, 545−550. (21) Aguilar-Martínez, R.; Palacios-Corvillo, M. A.; Greenwood, R.; Mills, G. A.; Vrana, B.; Gómez-Gómez, M. M. Calibration and use of the Chemcatcher® passive sampler for monitoring organotin compounds in water. Anal. Chim. Acta 2008, 618 (2), 157−167. (22) Schmitt-Jansen, M.; Altenburger, R. Community-level microalgal toxicity assessment by multiwavelength-excitation PAM fluorometry. Aquat. Toxicol. 2008, 86 (1), 49−58. (23) Schmitt-Jansen, M.; Altenburger, R. Predicting and observing responses of algal communities to photosystem II-herbicide exposure

ASSOCIATED CONTENT

S Supporting Information *

Additional information on chemical identity and characteristics of the studied compounds, physico-chemical water parameters of the sampling sites, time dependence of the effects, parameter of the concentration−response relationships and statistical tests. The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.est.5b01297.

■

AUTHOR INFORMATION

Corresponding Author

*Phone: +49 341 2351568; e-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank Iris Christmann and Klaus Seyfarth for technical assistance. This study was kindly supported by Helmholtz Impulse and Networking Fund through the Helmholtz Interdisciplinary Graduate School for Environmental Research (HIGRADE).

■

REFERENCES

(1) Schwarzenbach, R. P.; Escher, B. I.; Fenner, K.; Hofstetter, T. B.; Johnson, C. A.; von Gunten, U.; Wehrli, B. The challenge of micropollutants in aquatic systems. Science 2006, 313 (5790), 1072− 1077. (2) Brack, W. Effect-directed analysis: a promising tool for the identification of organic toxicants in complex mixtures? Anal. Bioanal. Chem. 2003, 377 (3), 397−407. (3) Brack, W.; Schmitt-Jansen, M.; Machala, M.; Brix, R.; Barcelo, D.; Schymanski, E.; Streck, G.; Schulze, T. How to confirm identified toxicants in effect-directed analysis. Anal. Bioanal. Chem. 2008, 390 (8), 1959−1973. (4) Mount, D. I.; Norberg-King, T. J.; Ankley, G. T.; Burkhardt, L. P.; Durhan, E. J.; Schubauer-Berigan, M. K.; Lukasewycz, M.; Blanck, H. Methods for Aquatic Toxicity Identification Evaluations. Phase III Toxicity Confirmation Procedures for Samples Exhibiting Acute and Chronic toxicity, Vol EPA/600/R-92/081; Environmental Research Laboratory, 10055

DOI: 10.1021/acs.est.5b01297 Environ. Sci. Technol. 2015, 49, 10048−10056

Article

Environmental Science & Technology using pollution-induced community tolerance and species-sensitivity distributions. Environ. Toxicol. Chem. 2005, 24 (2), 304−312. (24) Krammer, K.; Lange-Bertalot, H., Bacillariophyceae, Teil 1−5. In Süßwasserflora Mitteleuropa; Ettl. H., Gerloff, J., Heynig, H., Mollenhauer, D., Eds.; Gustav Fischer Verlag: Stuttgart: New York, 1986−1997. (25) Lecointe, C.; Coste, M.; Prygiel, J. “Omnidia”: software for taxonomy, calculation of diatom indices and inventories management. Hydrobiologia 1993, 269−270 (1), 509−513. (26) Ziemann, H.; Nolting, E.; Rustige, K. H., Bestimmung des Halobienindex - Methoden der biologischen Gewässeruntersuchung. In Band 2 Biologische Gewässeruntersuchung; von Tümpling, W., Friedrich, G., Eds.; Gustav Fischer: Jena, 1999. (27) Arrhenius, A.; Backhaus, T.; Gronvall, F.; Junghans, M.; Scholze, M.; Blanck, H. Effects of three antifouling agents on algal communities and algal reproduction: mixture toxicity studies with TBT, Irgarol, and Sea-Nine. Arch. Environ. Contam. Toxicol. 2006, 50 (3), 335−345. (28) Segner, H.; Schmitt-Jansen, M.; Sabater, S. Assessing the Impact of Multiple Stressors on Aquatic Biota: The Receptor’s Side Matters. Environ. Sci. Technol. 2014, 48 (14), 7690−7696. (29) Eloranta, P.; Soininen, J. Ecological status of some Finnish rivers evaluated using benthic diatom communities. J. Appl. Phycol. 2002, 14 (1), 1−7. (30) Schaumburg, J.; Schranz, C.; Foerster, J.; Gutowski, A.; Hofmann, G.; Meilinger, P.; Schneider, S.; Schmedtje, U. Ecological classification of macrophytes and phytobenthos for rivers in Germany according to the water framework directive. Limnologica 2004, 34, 283−301. (31) Jan Stevenson, R.; Glover, R. Effects of algal density and current on ion transport through periphyton communities. Limnol.Oceanogr. 1993, 38 (6), 1276−1281. (32) Ivorra, N.; Bremer, S.; Guasch, H.; Kraak, M. H. S.; Admiraal, W. Differences in the sensitivity of benthic microalgae to ZN and CD regarding biofilm development and exposure history. Environ. Toxicol. Chem. 2000, 19 (5), 1332−1339. (33) Faust, M.; Altenburger, R.; Backhaus, T.; Blanck, H.; Boedeker, W.; Gramatica, P.; Hamer, V.; Scholze, M.; Vighi, M.; Grimme, L. H. Predicting the joint algal toxicity of multi-component s-triazine mixtures at low-effect concentrations of individual toxicants. Aquat. Toxicol. 2001, 56 (1), 13−32. (34) Schäfer, R. B.; von der Ohe, P. C.; Rasmussen, J.; Kefford, B. J.; Beketov, M. A.; Schulz, R.; Liess, M. Thresholds for the effects of pesticides on invertebrate communities and leaf breakdown in stream ecosystems. Environ. Sci. Technol. 2012, 46 (9), 5134−5142. (35) European Commission. Common Implementation Strategy for the Water Framework Directive (2000/60/EC). Guidance Document No. 19guidance on Surface Water Chemical Monitoring under the Water Framework Directive, 2009. (36) Molander, S.; Dahl, B.; Blanck, H.; Jonsson, J.; Sjöström, M. Combined effects of tri-n-butyl tin (TBT) and diuron on marine periphyton communities detected as pollution-induced community tolerance. Arch. Environ. Contam. Toxicol. 1992, 22 (4), 419−427. (37) Sans-Piché, F.; Klünder, C.; Altenburger, R.; Schmitt-Jansen, M. Anchoring metabolic changes to phenotypic effects in the chlorophyte Scenedesmus vacuolatus under chemical exposure. Mar. Environ. Res. 2010, 69, 28−30. (38) Schwab, K.; Altenburger, R.; Lubcke-von Varel, U.; Streck, G.; Brack, W. Effect-directed analysis of sediment-associated algal toxicants at selected hot spots in the river Elbe basin with a special focus on bioaccessibility. Environ. Toxicol. Chem. 2009, 28 (7), 1506− 1517. (39) TCI America, T., Material safey data sheet: Phenyl-2naphthalene. 2012. https://www.spectrumchemical.com/MSDS/TCIP0198.pdf. (40) Blanck, H.; Dahl, B. Recovery of marine periphyton communities around a Swedish marina after the ban of TBT use in antifouling paint. Mar. Pollut. Bull. 1998, 36, 437−442.

(41) Blanck, H.; Wängberg, S.-Å. Pattern of cotolerance in marine periphyton communities established under arsenate stress. Aquat. Toxicol. 1991, 21, 1−14. (42) Kim Tiam, S.; Morin, S.; Pesce, S.; Feurtet-Mazel, A.; Moreira, A.; Gonzalez, P.; Mazzella, N. Environmental effects of realistic pesticide mixtures on natural biofilm communities with different exposure histories. Sci. Total Environ. 2014, 473−474, 496−506. (43) Pesce, S.; Morin, S.; Lissalde, S.; Montuelle, B.; Mazzella, N. Combining polar organic chemical integrative samplers (POCIS) with toxicity testing to evaluate pesticide mixture effects on natural phototrophic biofilms. Environ. Pollut. 2011, 159 (3), 735−741.

10056

DOI: 10.1021/acs.est.5b01297 Environ. Sci. Technol. 2015, 49, 10048−10056