Aquatic Toxicology 155 (2014) 348–359

Contents lists available at ScienceDirect

Aquatic Toxicology journal homepage: www.elsevier.com/locate/aquatox

Toxicity of lead (Pb) to freshwater green algae: Development and validation of a bioavailability model and inter-species sensitivity comparison K.A.C. De Schamphelaere ∗ , C. Nys, C.R. Janssen Faculty of Bioscience Engineering, Laboratory of Environmental Toxicology and Aquatic Ecology, Environmental Toxicology Unit (GhEnToxLab), Ghent University (UGent), Jozef Plateaustraat 22, B-9000 Gent, Belgium

a r t i c l e

i n f o

Article history: Received 18 February 2014 Received in revised form 4 July 2014 Accepted 8 July 2014 Available online 15 July 2014 Keywords: Bioavailability Fulvic acid Water hardness Biotic ligand model Risk assessment Water quality criteria Environmental quality standards

a b s t r a c t Scientifically sound risk assessment and derivation of environmental quality standards for lead (Pb) in the freshwater environment are hampered by insufficient data on chronic toxicity and bioavailability to unicellular green algae. Here, we first performed comparative chronic (72-h) toxicity tests with three algal species in medium at pH 6, containing 4 mg fulvic acid (FA)/L and containing organic phosphorous (P), i.e. glycerol-2-phosphate, instead of PO4 3− to prevent lead-phosphate mineral precipitation. Pseudokirchneriella subcapitata was 4-fold more sensitive to Pb than Chlorella kesslerii, with Chlamydomonas reinhardtii in the middle. The influence of medium physico-chemistry was therefore investigated in detail with P. subcapitata. In synthetic test media, higher concentrations of fulvic acid or lower pH protected against toxicity of (filtered) Pb to P. subcapitata, while effects of increased Ca or Mg on Pb toxicity were less clear. When toxicity was expressed on a free Pb2+ ion activity basis, a log-linear, 260-fold increase of toxicity was observed between pH 6.0 and 7.6. Effects of fulvic acid were calculated to be much more limited (1.9-fold) and were probably even non-existent (depending on the affinity constant for Pb binding to fulvic acid that was used for calculating speciation). A relatively simple bioavailability model, consisting of a log-linear pH effect on Pb2+ ion toxicity linked to the geochemical speciation model Visual Minteq (with the default NICA-Donnan description of metal and proton binding to fulvic acid), provided relatively accurate toxicity predictions. While toxicity of (filtered) Pb varied 13.7-fold across 14 different test media (including four Pb-spiked natural waters) with widely varying physico-chemistry (72h-EC50s between 26.6 and 364 ␮g/L), this bioavailability model displayed mean and maximum prediction errors of only 1.4 and 2.2-fold, respectively, thus indicating the potential usefulness of this bioavailability model to reduce uncertainty in site-specific risk assessment. A model-based comparison with other species indicated that the sensitivity difference between P. subcapitata and two of the most chronically Pbsensitive aquatic invertebrates (the crustacean Ceriodaphnia dubia and the snail Lymnaea stagnalis) is strongly pH dependent, with P. subcapitata becoming the most sensitive of the three at pH > 7.4. This indicates that inter-species differences in Pb bioavailability relationships should be accounted for in risk assessment and in the derivation of water quality criteria or environmental quality standards for Pb. The chronic toxicity data with three algae species and the bioavailability model presented here will help to provide a stronger scientific basis for evaluating ecological effects of Pb in the freshwater environment. © 2014 Published by Elsevier B.V.

1. Introduction Data about the chronic freshwater ecotoxicity of metals are important for risk assessment (RA) (e.g. REACH) and for the

∗ Corresponding author. Tel.: +32 9 264 37 64; fax: +32 9 264 37 66. E-mail addresses: [email protected] (K.A.C. De Schamphelaere), [email protected] (C. Nys), [email protected] (C.R. Janssen). http://dx.doi.org/10.1016/j.aquatox.2014.07.008 0166-445X/© 2014 Published by Elsevier B.V.

derivation of water quality criteria (WQC), such as an annualaverage environmental quality standard (AA-EQS) in the context of the European Water Framework Directive (EC, 2000, 2008a,b, 2011) or a criterion continuous concentration (CCC) in the United States of America (Stephan et al., 1985). Over the past decades, many chronic toxicity data for lead (Pb), based on measured Pb concentrations, have been reported for fish (e.g., Holcombe et al., 1976; Burden et al., 1998; Grosell et al., 2006a) and invertebrates (e.g., Besser et al., 2005; Grosell et al., 2006b; Mager et al., 2011;

K.A.C. De Schamphelaere et al. / Aquatic Toxicology 155 (2014) 348–359

Esbaugh et al., 2012). In contrast, much fewer data are available for a third important taxonomic group, i.e. the green algae. Although a considerable number of studies have investigated biochemical effects of Pb on algal species (e.g., reactive oxygen species production, Szivak et al., 2009), relatively few studies are available that have reported effects on the basis of measured Pb concentrations for endpoints such as growth rate or cell or biomass yield, which are the algal endpoints that are typically considered relevant for regulatory use (EC, 2011; Stephan et al., 1985). A recent search of U.S. EPA’s AQUIRE database for such studies only returned one study (Stokes, 1981) that reported Pb toxicity to a green alga (Scenedesmus acuminatus) on the basis of a measured Pb concentration (with an EC50 for cell growth of 250 ␮g/L of total Pb), but their exposure duration (6–8 days) was not the standard 3 days as is currently prescribed in internationally accepted test guidelines for algae (OECD, 2011). Therefore, the first aim of our study was to investigate chronic Pb toxicity using standard 3d-toxicity tests with three commonly tested algal species, i.e. Pseudokirchneriella subcapitata, Chlorella kesslerii, and Chlamydomonas reinhardtii. This would not only provide chronic ecotoxicity data for green algae that are in line with currently accepted regulatory standards, but would also provide knowledge about the interspecies variability of Pb sensitivity within this taxonomic group. Furthermore, it is currently generally accepted that RA and WQC derivation for metals should take into account the influence of water chemistry variables (such the concentration of dissolved organic carbon (DOC), pH, and water hardness) on toxicity by means of bioavailability models like the biotic ligand model (BLM) (Van Sprang et al., 2009; ECI, 2008; EC, 2008a,b; OECD, 2012; Schlekat et al., 2010). Bioavailability of Pb has been investigated with considerable detail for fish and cladocerans, pointing to important effects of pH and DOC (fish and cladocerans, Grosell et al., 2006a and Mager et al., 2011) and Ca (only fish, Grosell et al., 2006a) on chronic toxicity of Pb, and validated bioavailability models are currently available for Ceriodaphnia dubia (Esbaugh et al., 2012; Nys et al., 2014) and Lymnaea stagnalis (Esbaugh et al., 2012). The same variables (pH, DOC and Ca) have also been shown to influence uptake of Pb into green algae (Slaveykova and Wilkinson, 2002, 2003; Slaveykova et al., 2003), but studies about their effect on Pb induced algal growth inhibition are very scarce. One study showed an influence of pH and Ca (but not of humic acid concentration) on Pb induced algal growth inhibition (Pawlik-Skowronska, 2002), but because Pb concentrations were not measured during exposures, we considered those data insufficiently reliable for bioavailability model development. Therefore, the second aim of our study was to investigate the effect of DOC (added as Suwannee River Fulvic Acid, SRFA), pH, Ca, and Mg on chronic Pb toxicity to the most sensitive of the above-mentioned three species (i.e. P. subcapitata, see Results and Discussion, section 3.1) again using standard 3-day growth inhibition assays. Toxicity of Pb is reported as EC10s, EC20s and EC50s, expressed on the basis of measured filtered (84% of total Pb in solution (Supplementary Material, Table S1.2). In contrast, in standard OECD test medium (at pH 7.6) only 9.6% of the total Pb was found back in the filtrate, meaning that >90% had precipitated in the form of particles >0.45 ␮m, with an apparent solubility limit of Pb between 100 and 200 ␮g/L (Supplementary Material, Table S1.1, Figure S1.1). The relation between algal growth inhibition and measured filtered Pb in this medium was also non-monotonous (Supplementary Material, Figure S1.2). Together, these data confirm the inappropriateness of media containing high inorganic PO4 concentration for the generation of reliable ecotoxicity data for regulatory purposes (Kopittke et al., 2008). These data also indicate that replacement of PO4 by Gly-2-P in OECD medium provides a valuable alternative medium for Pb toxicity testing with algae. This was further confirmed in our subsequent tests in synthetic media, where measured initial filtered Pb concentrations were always between 80 and 96% of the measured total Pb concentration (see Results and Discussion, section 3.2.1) and by the fact that Gly-2-P addition did not seem to affect speciation of Pb2+ , as shown by speciation measurements and calculations (see Material and Methods, section 2.3.2; see Supplementary Material 2). 2.3.2. Test media phase 2—Bioavailability of Pb to P. subcapitata This test phase consisted of three test series. All test series consisted of 3 or 4 toxicity tests that were run simultaneously in 3 or 4 test media with different chemistry (Table 1). In these test series the univariate effect of increasing DOC (added as SRFA; series 1), Ca or Mg (added as CaCl2 or MgCl2 , series 2) and pH (series 3) on Pb toxicity to P. subcapitata were investigated.

The basis growth medium for all tests in test phase 2 was OECD medium (OECD, 2011), with following modifications. First, as above (section 2.3.1), EDTA was omitted and inorganic phosphate was replaced with Gly-2-P. Second, 10 mM MES (pH 6) or 3.6 mM 3(N-morpholino)propanesulfonic acid (MOPS: pH > 6) was added as a pH buffer. Third, appropriate amounts of SRFA, CaCl2 , MgCl2 , NaOH and/or HCl were added to reach the desired target levels of SRFA, Ca, Mg and pH (see Table 1 for target levels of these variables, see Table 2 for detailed chemistry of test media during tests). As a further support of our choice to replace inorganic phosphate with Gly-2-P (see section 2.3.1), we confirmed with actual speciation measurements (using the Wageningen Donnan Membrane Technique, DMT; Temminghoff et al., 2000; Weng et al., 2005, 2010) in one of these test media (i.e., the 3 mM Ca medium) that addition of Gly-2-P did not affect free Pb2+ ion activity compared to the same medium without any P-addition (Supplementary Material 2.1). Furthermore, speciation calculations for all test media used strongly indicate that there is no influence of adding glycerol-2-phosphate to the test medium on free Pb2+ ion activities (Supplementary Material 2.2). 2.3.3. Test media phase 3—Ecotoxicity of Pb to P. subcapitata in Pb-spiked surface waters Ten liter of field water was collected at four locations (Table 2), i.e. Ankeveensche Plas (a Ditch in Ankeveen, the Netherlands, denoted as Ankeveen), Markermeer (a large lake in Marken, the Netherlands, denoted as Markermeer), Beneden Regge (a river, Ommen, the Netherlands, denoted as Regge), and Ruisseau de Saint Martin (a small stream, Bihain, Belgium, denoted as Bihain). Samples were filtered through a 0.45 ␮m filter before further analysis and testing. Samples were taken for measurement of DOC, major ions, PO4 3− and trace metals (Pb, Fe, Al, Zn, Cu, Ni, Cd). A 2L aliquot was then prepared for algae testing by adding all components of the OECD TG201 medium (OECD, 2011), except CaCl2 , MgSO4 , KH2 PO4 , Na2 EDTA·2H2 O and NaHCO3 . MOPS buffer (3.6 mM) was added to Ankeveen and Regge; MES buffer (10 mM) was added to Bihain; no pH buffer was added to Markermeer. The pH was adjusted to pH measured in the field using dilute NaOH. The chemical composition during testing of all waters is given in Table 2. 2.3.4. Timing of media preparation and Pb-spiking Media (without Pb) were always prepared four days prior to the start of the test and equilibrated with atmosphere at testing temperature (24 ◦ C). Pb was spiked into the media 24 h prior to the start of the test (i.e. 24 h prior to inoculation with the algal cells). 2.4. Chemical analyses during the tests DOC and dissolved inorganic carbon (DIC) were measured using a TOC analyzer (TOC-5000, Shimadzu, Duisburg, Germany). The measured DOC concentration in the synthetic waters with SRFA additions was always within 15% of the nominal added

K.A.C. De Schamphelaere et al. / Aquatic Toxicology 155 (2014) 348–359

351

Table 2 Detailed chemical compositiona of all synthetic media and natural waters used for investigating Pb toxicity to P. subcapitata. This chemical composition was used as input for speciation modeling with Visual Minteq. Test medium

MOPS (mM)

MES (mM)

pH

FAb (mg/L)

DOCb (mg/L)

Ca (mM)

Mg (mM)

Na (mM)

K (mM)

Cl (mM)

SO4 (mM)

DIC (mM)

1 2 3

FA 4 mg/L FA 10 mg/L FA 20 mg/L

0 0 0

10 10 10

6.0 6.0 6.0

4.0 10.0 20.0

2.1 5.3 10.6

0.122 0.122 0.122

0.120 0.120 0.120

3.595 3.695 3.695

0.009 0.009 0.009

0.642 0.642 0.642

0.0609 0.0609 0.0609

0.083 0.083 0.083

4 5 6 7

Ca 0.12 mM, Mg 0.12 mM Ca 1 mM, Mg 0.12 mM Ca 3 mM, Mg 0.12 mM Ca 0.12 mM, Mg 1 mM

3.6 3.6 3.6 3.6

0 0 0 0

7.0 7.0 7.0 7.0

4.0 4.0 4.0 4.0

2.1 2.1 2.1 2.1

0.122 1.000 3.000 0.122

0.120 0.120 0.120 1.000

1.645 1.595 1.595 1.545

0.009 0.009 0.009 0.009

0.642 2.398 6.398 2.402

0.0609 0.0609 0.0609 0.0609

0.383 0.383 0.383 0.358

8 9 10

pH 6.0 pH 6.8 pH 7.6

0 3.6 3.6

10 0 0

6.0 6.8 7.6

4.0 4.0 4.0

2.1 2.1 2.1

0.122 0.122 0.122

0.120 0.120 0.120

3.545 1.445 3.195

0.009 0.009 0.009

0.642 0.642 0.642

0.0609 0.0609 0.0609

0.083 0.292 0.675

11 12 13 14

Markermeer Ankeveen Regge Bihain

0 3.6 3.6 0

0 0 0 10

8.0 7.0 7.7 6.0

8.5 29.1 20.7 8.5

6.5 22.4 15.9 6.5

1.373 2.046 1.455 0.118

0.535 0.447 0.290 0.069

7.370 3.230 5.200 3.401

0.244 0.033 0.228 0.020

6.292 1.678 3.882 0.574

0.783 2.453 0.104 0.104

3.134 0.559 4.226 0.049

Test ID

a MOPS = 3-N-morpholino-propane-sulfonic acid, MES = 2-(N-morpholino)-ethane-sulfonic acid, FA = fulvic acid, DOC = dissolved organic carbon, DIC = dissolved inorganic carbon. b For synthetic media (ID 1–10) DOC = 0.5244 × FA; for natural waters (ID 11–14) we assumed FA = 1.3 × DOC (see Materials and Methods for details).

concentration of DOC and therefore the nominal concentrations were used for modeling. In the four natural surface waters, major cations (Ca, Mg, Na, and K) were measured with ICP-OES (Perkin Elmer 3300DV); major anions (Cl, SO4 ) were measured with colorimetry. Samples of total and filtered Pb (0.45 ␮m Gelman Sciences, Ann Arbor, MI, USA) were acidified (1% v/v) with 0.14 N HNO3 (NormatomTM , ultrapure, VWR, Leuven, Belgium) and were measured by flame-AAS (SpectrAA100, Varian, Mulgrave, Australia; for nominal concentrations >200 ␮g Pb/L) or ICP-MS (for nominal concentrations 35%. However, as all validity criteria were fulfilled up to 2 days of exposure, the 2d-ECx values were calculated for these two species as an alternative (Table 3). For P. subcapitata, the 2d-ECx values were very similar to the 3dECx values, i.e. within 10% of each other and with overlapping 95% confidence intervals. A reliable comparison of species sensitivity was therefore possible on the basis of 2d-growth rates. Based on the concentration response data and the fitted log-logistic concentration response curves, it is clear that P. subcapitata is the most sensitive and C. kesslerii the least sensitive, with C. reinhardtii in the middle (Table 3, Fig. 1). P. subcapitata has 2d-EC10Pbfilt , 2d-EC20Pbfilt and 2d-EC50Pbfilt values of 29.7, 44.7, and 89.9 ␮g/L, respectively; which are 4-fold lower than those of C. kesslerii. P. subcapitata has been shown to be the most sensitive green algae species to other metals in some other inter-species sensitivity comparison studies (e.g., De Schamphelaere and Janssen, 2006; Van Sprang et al., 2009), but not in others (e.g., Lee et al., 2005; Deleebeeck et al., 2009b). In the present study, we chose to use the most sensitive species to Pb, i.e. P. subcapitata, as the model algae for investigating Pb bioavailability, i.e. in the investigation of how Pb toxicity is influenced by physico-chemistry (see section 3.2).

352

K.A.C. De Schamphelaere et al. / Aquatic Toxicology 155 (2014) 348–359

Table 3 Results of the chronic toxicity tests with lead (Pb) with three species of green algae in modified OECD mediuma for interspecies comparison, including details on test validity criteriab and effect concentrations (expressed as ␮g filtered Pb/L). The 95% confidence intervals are between parentheses. Species

Exposure Time (days)

Growth rate in control (d−1 )

P. subcapitata P. subcapitata C. kesslerii C. kesslerii C. reinhardtii C. reinhardtii

2 3 2 3 2 3

1.451 1.303 1.388 1.085 1.383 1.042

Factor increase of cell density in control

CV of control growth rate (%)

18.3 50.3 16.1 24.4 16.0 22.4

4.0 2.0 1.6 5.7 4.8 2.2

CV of sectional growth rate in control (%)

EC50Pbfilt (␮g/L)

EC20Pbfilt (␮g/L)

EC10Pbfilt (␮g/L)

16.3 23.2 7.0 54.1 29.0 64.3

89.9 (77.2–104.7) 83.9 (72.3–97.4) 388 (342–440) NDc 172 (155–190) NDc

44.7 (34.8–57.3) 45.7 (35.7–58.5) 185 (148–231) NDc 108 (89–130) NDc

29.7 (21.2–41.4) 32.0 (22.9–44.8) 120 (86–166) NDc 82.3 (63.5–106.6) NDc

a

See main text for details; chemical composition of medium is the same as the medium with test ID1 (see Table 2), except for the Na concentration (see main text). Validity criteria based on control performance (OECD, 2011): mean growth rate c > 0.92, mean factor cell increase FIc > 16, Coefficient of Variation (CV) of control growth rate CV␮,c 0.92, mean factor cell increase FIc > 16, Coefficient of Variation (CV) of control growth rate CV␮,c < 7% for P. subcapitata and

Toxicity of lead (Pb) to freshwater green algae: development and validation of a bioavailability model and inter-species sensitivity comparison.

Scientifically sound risk assessment and derivation of environmental quality standards for lead (Pb) in the freshwater environment are hampered by ins...
944KB Sizes 0 Downloads 3 Views