Multiwalled Carbon Nanotubes at Environmentally Relevant

Subscriber access provided by EASTERN KENTUCKY UNIVERSITY

Article

Multiwalled carbon nanotubes at environmentally relevant concentrations affect the composition of benthic communities Ilona Velzeboer, Edwin Peeters, and Albert Aart Koelmans Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/es400777j • Publication Date (Web): 28 May 2013 Downloaded from http://pubs.acs.org on June 3, 2013

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Environmental Science & Technology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 28

Environmental Science & Technology 1

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

Multiwalled carbon nanotubes at environmentally relevant concentrations affect the

2

composition of benthic communities

3

I. Velzeboer*,†,‡, E.T.H.M. Peeters†, A.A. Koelmans†,‡

4

† Wageningen University, Aquatic Ecology and Water Quality Management Group, P.O. Box

5

47, 6700 AA Wageningen, The Netherlands

6

‡ IMARES – Institute for Marine Resources & Ecosystem Studies, Wageningen UR, P.O. Box

7

68, 1970 AB IJmuiden, The Netherlands

8 9

* Corresponding author; e-mail: [email protected], tel: +31 317 48 71 44

10

ACS Paragon Plus Environment

Environmental Science & Technology

Page 2 of 28 2

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

11

Abstract

12

To date, chronic effect studies with manufactured nanomaterials under field conditions are

13

scarce. Here, we report in situ effects of 0, 0.002, 0.02, 0.2 and 2 g/kg multiwalled carbon

14

nanotubes (MWCNTs) in sediment on the benthic community composition after 15 months of

15

exposure. Effects observed after 15 months were compared to those observed after 3 months and

16

to community effects of another carbonaceous material (activated carbon; AC) which was

17

simultaneously tested in a parallel study. Redundancy analysis with variance partitioning

18

revealed a total explained variance of 51.7% of the variation in community composition after 15

19

months, of which MWCNT dose explained a statistically significant 9.9% . By stepwise

20

excluding the highest MWCNT concentrations in the statistical analyses, MWCNT effects were

21

shown to be statistically significant already at the lowest dose investigated, which can be

22

considered environmentally relevant. We conclude that despite prolonged aging, encapsulation

23

and burial, MWCNTs can affect the structure of natural benthic communities in the field. This

24

effect was similar to that of AC observed in a parallel experiment, which however was applied at

25

a fifty times higher maximum dose. This suggests that the benthic community was more

26

sensitive to MWCNTs than to the bulk carbon material AC.

27 28


Page 3 of 28


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

29

Introduction

30

In the last years, there has been an increase in efforts to assess the fate and effects of engineered

31

nanoparticles (ENPs) in the environment.1, 2 These efforts have largely focused on nanoparticles

32

in water but have also increasingly included aquatic sediments, which are argued to be an

33

important sink for nanoparticles.3-6 For traditional pollutants, effect assessment has evolved from

34

single species standard toxicity testing on laboratory scale to chronic and in situ effect studies,

35

including endpoints on the population and community level.7 For ENPs, however, assessment of

36

in situ effects still is in its infancy. Most ENP ecotoxicity studies focus on the development of

37

single species laboratory tests and use ENP doses far above environmentally relevant

38

concentrations.2-4, 8-10 Such tests are crucial to derive dose-response relationships in the

39

framework of hazard assessment, and to identify mechanisms of toxicity on the species level.2

40

However, they are less realistic with respect to multiple exposure routes, dynamic particle

41

mixing, effects fouling and aging of ENPs, or ecological processes on the community level like

42

competition, predation, reproduction or recolonization. Alterations of ENPs on the particle scale

43

depend on the exposure pathway and environmental history of the particles in the field and are

44

likely to alter their ultimate behaviour and toxicity.11 Timescales for the expression of ENP

45

toxicity thus may exceed those of standard laboratory tests and may include species-specific sub-

46

lethal or behavioural endpoints, which may only become visible in long term multi-generation

47

community assessments.12 Consequently, assessment of in situ effects is urgently needed because

48

laboratory studies may be expected to be poor predictors of effects under field conditions.1 This

49

may be true especially for ENPs for the reasons stated, but also because exposure assessment in

50

current ENP toxicity tests is much less well controlled than for traditional chemicals.2



Page 4 of 28 4

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

51

The aim of the present study was to assess long-term macroinvertebrate community effects of

52

multiwalled carbon nanotubes (MWCNTs) in sediment, including doses that can be considered

53

environmentally relevant. MWCNTs constitute an important category of the carbon-based

54

nanoparticles, used in all kinds of products,13 and suggested as a sorbent for the remediation

55

treatment of contaminated sediments or waste waters.13-15 Like all carbon nanotubes, MWCNTs

56

are mostly hydrophobic and have been shown to aggregate with organic matter, other

57

nanoparticles or suspended solids in aquatic systems,13 after which they are rapidly removed

58

from the water column by sedimentation.5, 16, 17 Consequently, exposure of benthic organisms to

59

MWCNTs is plausible. Whereas previous work addressed community effects after relatively

60

short exposure times of 3 months,18 the present study focussed on the influence of MWCNT-

61

contaminated sediment on the recolonization of benthic communities in a controlled and

62

replicated field experiment that lasted 15 months. MWCNT treatment effects after 15 months

63

were also quantitatively compared to community composition after 3 months. Furthermore, these

64

MWCNT effects were compared to community effects caused by another carbonaceous material

65

(activated carbon; AC) which was applied simultaneously in parallel systems in the framework

66

of another study.19 By combining the current data on 15 month exposure to MWCNTs with those

67

of 3 month exposure and those of the AC treatments, the effects of exposure time and carbon

68

type could be assessed, in addition to MWCNT effects alone.

69 70

Materials and methods

71

Community experiment. The recolonization field experiment followed previously published

72

procedures.18, 19 Briefly, the experiment consisted of collecting sediment from the experimental

73

site, elimination of invertebrates from the sediments, thorough mixing of MWCNTs into the


Page 5 of 28


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

74

sediments, transferring the MWCNT-amended sediments into trays, placing the trays with

75

treated sediment back into the original site and monitoring the trays for colonization of species

76

originating from the surrounding (continuous) donor system. The field experiment was

77

performed in an uncontaminated ditch (2-3 m width, 1.5 m depth) at the “Veenkampen”, which

78

is an experimental field site of Wageningen University in the Netherlands (51°58'52"N,

79

5°37'25"E). Organic contaminants, total organic carbon (TOC) and native black carbon (BC)

80

concentrations in the sediment were measured in the framework of an earlier study,19 and are

81

provided as Supporting Information (Table S1). Additionally, organic matter (OM) content was

82

determined as loss on ignition (3 h, 550°C) and was 20% (SD 0.24). Prior to exposure, the

83

benthic macroinvertebrate community of the study site was investigated and contained mostly

84

Oligochaeta, Hirundinea, Gastropoda, Bivalvia, Arachnida, Crustacea, and Insecta.

85

In May 2009, the upper 5 to 10 cm of the sediment in an area of 3 by 10 m was sampled using a

86

0.5 mm mesh size net. The sediment was sieved with a 2 mm mesh stainless steel screen to

87

remove pebbles and large organic debris. Macroinvertebrates were eliminated from the sediment

88

by creating anoxic conditions to prevent false colonization during the experiment. This was done

89

by storing the sediment for 2 weeks at 20°C in closed 26 L polyethylene barrels without

90

headspace. Powdered MWCNTs with an inner diameter of 5-10 nm, outer diameter of 20-30 nm,

91

length of 10-30 μm and a purity of 95 wt.% were obtained from Cheaptubes (Brattleboro VT,

92

USA). Details on metal concentrations and on MWCNT characterisation by transmission

93

electron microscopy (TEM images) are provided as Supporting Information (Table S2 and

94

Figure S1). Sediment concentrations of MWCNT associated metals were far below toxicity

95

thresholds and thus were too low to cause effects on the community (see Table S2). The

96

MWCNTs were added to the sediment as a powder and thoroughly mixed on a roller bank for 6



Page 6 of 28 6

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

97

hours to obtain final concentrations of 0, 0.002, 0.02, 0.2 and 2 g/kg dry weight as described in

98

Velzeboer et al.18 This range covers low but environmentally realistic concentrations anticipated

99

from fate model predictions,5, 20, 21 as well as higher levels at which actual effects have been

100

observed in invertebrate single species laboratory tests.5, 13, 22-27 Open polypropylene trays

101

(30x15.5x20 cm) were filled with 2 L MWCNT treated sediment each, to obtain 4 replicates per

102

MWCNT concentration and 5 replicates for the control. The trays were embedded randomly in a

103

preselected grid in the unpolluted donor system where the sediment originally came from. The

104

trays were carefully filled with water from the ditch before inserting them 15 cm into the donor

105

sediment. The top of the tray walls emerged 5 cm above the sediment level to prevent

106

surrounding sediment from entering the trays. Previous work has shown that the tray walls did

107

not prevent recolonization for any of the species originally present.18, 19 To prevent the trays from

108

sinking too far in the sediment, they were connected to wires spanning the surface of the ditch.

109

Exposure lasted for 15 months (June 2009 till October 2010), after which the trays were carefully

110

retrieved. Similarly, samples had been taken after 3 months for a short term MWCNT

111

community effect study in the same system.18 After retrieval, representative sediment

112

subsamples (30mL) were taken from the trays, combined per treatment and analyzed for organic

113

carbon (OC) and residual carbon (RC). OC and RC were measured by chemothermal oxidation

114

(CTO-375 method)28, 29 using a CHN analyzer (EA 110 CHN Elemental Analyzer, CE

115

Instruments, Milan, Italy). To improve comparability of the 3 months and 15 months OC and RC

116

data, original samples taken after 3 months and preserved by cooling, were reanalysed for OC

117

and RC.

118

Macroinvertebrates and macrophytes were collected from the trays using a 0.5 mm mesh sieve

119

and rinsing the sediment with tap water. Macrophytes were identified and dry weight was


Page 7 of 28


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

120

measured. Macroinvertebrates were preserved in 80% ethanol and stored at 4°C until

121

identification to genus or family level using available keys.30-32

122

At time zero and after 15 months, 0.3 m2 samples from the donor system outside the trays were

123

taken with a standard dipnet (mesh size 0.5 mm) for determination of the community

124

composition, OC and RC. To eliminate any chance of cross-contamination with the treated trays,

125

the donor system samples were taken at a safe distance of at least 10 m from the MWCNT-

126

treated trays.

127

Data analysis. Data analysis followed previously published procedures18, 19 and focused on (a)

128

long term MWCNT effects (15 months of exposure), (b) the effect of recolonization time, i.e.

129

differences between 0, 3 and 15 months post treatment, and (c) differences between MWCNT

130

and AC community effects. Data for 3 month exposure18 were reanalysed and original data for 3

131

and 15 months of exposure of AC were kindly provided by Kupryianchyk et al.19 Previous 3

132

month MWCNT and AC data18, 19 and current 15 month MWCNT data were obtained from

133

simultaneous exposures in the same donor system, and therefore are directly comparable. Effects

134

of MWCNTs on the benthic community were quantified using univariate as well as multivariate

135

community measures. Univariate endpoints were number of individuals, number of taxa and the

136

Shannon diversity index (H), describing the appearance as well as the homogeneity of the

137

variance within the community: ∑(

)

138

where pi is the fraction of the total number of individuals of taxon i and n is the number of taxa

139

in the community.33 Normality of the data was tested using the Shapiro-Wilk test. Statistical

140

significance of effects was tested using one-way ANOVA with Tukey’s Post Hoc test. To test the

141

differences of the univariate endpoints between 3 and 15 months of exposure, an independent t



Page 8 of 28 8

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

142

test was used for each treatment. All analyses were performed using SPSS (IBM SPSS Statistics

143

19.0, IBM, Amonk, NY, USA) with p < 0.05 as criterion for significance.

144

Multivariate analysis included redundancy analysis (RDA) with partitioning of the variance to

145

detect patterns and origins of variation in the macroinvertebrate data.19, 34-38 Prior to multivariate

146

analysis, the data were Log(abundance+1) transformed. First, RDA was performed to quantify

147

the variance explained by the explanatory variables MWCNT concentration, duration of

148

exposure, macrophyte biomass and spatial distribution in the ditch together. Spatial distribution

149

was expressed as coordinates, by numbering columns and rows in which the trays are placed.

150

Then, the explanatory power and significance of each variable was determined through a series

151

of single constrained RDAs,34-37 followed by Monte Carlo permutation tests with 499

152

permutations, for testing statistical significance.39 These analyses provide percentages of

153

variances explained by each of the explanatory variables and the significance level for each of

154

the percentages. Furthermore, the percentage of variance that remains unexplained was

155

calculated. The multivariate analyses were performed with CANOCO 4.5 for Windows.40

156 157

Results and discussion

158

System characteristics. During exposure there was no visible flow in the ditch. The water level

159

was regulated, so the trays maintained under water throughout the experiment. There were no

160

extreme weather conditions, i.e. wind-induced resuspension or other incidents that could have

161

interfered with the MWCNT level in the trays. OC after 3 and 15 months of exposure varied

162

slightly among treatments; from 7.8 ± 2.6 to 10.8 ± 2.8% d.w. and from 6.4 ± 1.4 to 9.4 ± 1.9 %

163

d.w. respectively (Figure 1a), thus showing some random variability between the treatments as

164

well as between time points. Similarly, RC content in the trays ranged from 0.41 ± 0.02 to 0.61 ±


Page 9 of 28


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

165

0.15% d.w. after 3 months of exposure, and from 0.21 ± 0.01 to 0.39 ± 0.04 % d.w., after 15

166

months of exposure (Figure 1b). Note that the presented errors relate to analytical error of the

167

OC and RC determinations on the mixed samples, not to variation among replicate systems in the

168

field. The observed RC percentages may be explained from the presence of natural kerogens or

169

natural BC.41 Theoretically, MWCNTs may also contribute to RC because carbon nanotubes

170

have been reported to partly contribute to RC as measured by chemothermal oxidation.42 The

171

highest nominal MWCNT dose of 0.2% comes close to the measured 0.24% RC after 15 months

172

(Figure 1b). However, the MWCNTs cannot explain the higher RC percentage at the much lower

173

MWCNT doses, which therefore have to relate to background natural kerogens or BC. The same

174

background kerogen and/or BC concentrations would also occur in the 0.2% MWCNT dose.

175

Therefore, it is unlikely that MWCNTs contributed substantially to RC in any of the treatments.

176

Overall, the average RC content per treatment decreased in time. The observed differences are

177

not to be considered large, given the natural temporal variability of sediment composition. We

178

have no conclusive explanation for the decrease, but this may relate to local inhomogeneities in

179

the sediment or local dilution with fresh organic matter. A similar dilution after 15 months was

180

observed simultaneously for AC at the same study site,19 which may imply a common

181

mechanism.

182

Univariate analysis of community effects. After 15 months, approximately all taxa returned in

183

their original abundance in the control trays (Table S4, Table S5), which shows that 15 months is

184

long enough for this benthic community to recover. Our previously published data after 3

185

months, showed that only 30% of the individuals and 50% of the taxa had returned in the control

186

trays.18 This indicates that in 15 months, full recovery of the community in terms of individuals

187

and taxa was achieved. We explain the difference between the time points from the way natural



Page 10 of 28 10

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

188

recolonization occurs. After 3 months, recolonization probably was dominated by active or

189

passive migration of individuals from the donor system to the trays, whereas after 15 months not

190

only migration but also reproduction contributed to the observed community composition.43

191

After 15 months a total of 26 different taxa were identified, ranging from 5 to 12 taxa per

192

tray (Figure S2). This is comparable to the previous results for 3 months with 32 taxa ranging

193

from 3 to 14 taxa per tray.18 After 15 months, no effect of the MWCNT treatment on the number

194

of taxa was detected (one-way ANOVA, F(4,16) = 1.341, p = 0.298; Figure S2), a lack of effect

195

also observed in the data after 3 months (one-way ANOVA, F(4,16) = 2.914, p = 0.05518). The

196

number of taxa increased between the 3 and 15 month exposure period for the control and the

197

lowest dose of 0.002 g/kg MWCNTs (independent t-test; 0 g/kg: t = -4.417, p = 0.002; 0.002

198

g/kg: t = -5.058, p = 0.002).

199

The number of individuals was 638-5236 per m2 after 15 months, compared to a much

200

lower 154-1540 per m2 after 3 months (Figure S3). Therefore, total abundance increased between

201

3 and 15 month exposure, which was statistically significant for 3 levels of MWCNTs

202

(independent t-test; 0 g/kg: t = -4.746, p = 0.010; 0.002 g/kg: t = -3.545, p = 0.012; 2 g/kg: t = -

203

2.443, p = 0.050). After 3 months, the number of individuals increased with increasing MWCNT

204

dose (one-way ANOVA, F(4,16) = 6.737, p = 0.002),18 but this was not the case after 15 months

205

(one-way ANOVA, F(4,16) = 0.951, p = 0.461). This may be explained by assuming that after

206

15 months a new stable situation was reached, whereas after 3 months the systems were not fully

207

recolonized, which may cause larger differences between the systems.

208

Biodiversity in the trays as quantified with the Shannon index increased slightly in time

209

from on average 1.28 ± 0.42 (3 months)18 to 1.58 ± 0.20 (15 months) (Figure S4). For both time

210

points, differences among MWCNT treatments were not statistically different (3 months: one-


Page 11 of 28


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

211

way ANOVA, F(4,16) = 2.204, p = 0.115;18 15 months: one-way ANOVA, F(4,16) = 1.035, p =

212

0.420). The Shannon indices between the time points increased significantly only for the 0.2 g/kg

213

MWCNT level (independent t-test; t = -3.686, p = 0.010).

214

Most abundant taxa in the 15 month systems were Erpobdella, Chironomidae,

215

Lumbriculidae and Sphaeriidae. They were also the most abundant taxa in the 3 month systems,

216

except for Chironomidae. In general, these species are also recognized to be less sensitive for

217

pollution.30, 34 The differences in absolute abundance between the MWCNT treatments were not

218

statistically significant for these dominant species after 3 nor after 15 months of exposure (one-

219

way ANOVA), except for Sphaeriidae abundance, which increased with MWCNT concentration

220

after 3 months (one-way ANOVA, F(4,16) = 7.287, p = 0.020). For Erpobdella, the higher

221

abundance after 15 months compared to the 3 month exposure was statistically significant for the

222

0 and 2 g/kg MWCNT levels (independent t-test; 0 g/kg: t = -3.859, p = 0.005; 2 g/kg: t = -2.988,

223

p = 0.024). For Chironomidae, the increase of the absolute abundance was shown to be

224

statistically significant for all MWCNT treatments after 15 months, compared to 3 months. The

225

abundance of Lumbriculidae was higher after 15 than after 3 months exposure, which was

226

statistically significant for the 0 and 0.2 g/kg MWCNT concentrations (independent t-test; 0

227

g/kg: t = -2.806, p = 0.023; 0.2 g/kg: t = -3.693, p = 0.010). Lumbriculidae thus developed over

228

time, which contrasts to results from simultaneously exposed parallel trays amended with up to

229

10% AC, where the abundance of this taxon decreased.19 The difference is probably explained by

230

the much higher AC dose used. For Sphaeriidae only the 0.2 g/kg MWCNT concentration was

231

significantly lower after 15 months compared to the 3 months exposure (independent t-test; 0.2

232

g/kg: t = 2.691, p = 0.036). In contrast to the other dominant species, the abundance of

233

Sphaeriidae decreased in time. This could be a direct negative effect of MWCNTs in time, but



Page 12 of 28 12

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

234

also the macrophyte abundance after 15 months was lower. This difference in habitat could also

235

play a role in the abundance of Sphaeriidae. The abundance of Sphaeriidae can be compared to

236

Pisidiidae in the parallel AC treatments,19 as both taxa are small freshwater clams from the same

237

Sphaeriidae family. For all doses of 0, 2, 4, 6 and 10% AC,19 the abundance of these clams

238

decreased significantly between 3 and 15 months, whereas for the MWCNTs, the decrease was

239

only significant for the 0.2 g/kg dose.

240

After 15 months, macrophyte abundance varied considerably among MWCNT treatments

241

and replicates, ranging from 0.68 ± 0.71 to 2.33 ± 2.15 g d.w. (Figure S5). The differences

242

between treatments were not statistically significant (one-way ANOVA, F(4,16) = 0.950, p =

243

0.461). Macrophyte taxa included Chara sp, Elodea nuttallii, Potamogeton obtusifulius, Alisma

244

plantago-aquatica and Fontinalis sp and one tray contained Eleocharis palustris. Interestingly,

245

the number of macroinvertebrate individuals increased with increasing macrophyte abundance

246

(Pearson’s correlation, r(21) = 0.550, p = 0.010) as could be expected since macrophytes offer

247

food, habitat and places to shelter for a variety of macroinvertebrate taxa. After 3 months, the

248

macrophyte abundance was reported to increase with increasing MWCNT concentration (one-

249

way ANOVA, F(4,16) = 4.394, p = 0.014; Figure S4).18 The difference of the macrophyte

250

abundance between 3 and 15 months was not statistically significant. This is similar to results

251

reported by Kupryanchyk et al.,19 showing that the variation in macrophyte abundance between

252

AC treatments was significant after 3 months but not anymore after 15 months.

253

Multivariate analysis of community data. Data from both time points and all MWCNT

254

treatments were analysed together to assess the effects of MWCNTs and other potential

255

explanatory variables on the community composition. Detrended correspondence analysis (DCA)

256

showed a gradient length of 2.024, indicating that a linear response model fitted best to this


Page 13 of 28


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

257

dataset. 44 RDA with partitioning of the variance showed that 51.7% of the variation (p = 0.002)

258

in the dataset was explained by the variables MWCNT dose, exposure time, macrophyte biomass

259

and spatial distribution (Table 1). General sediment characteristics like OC and RC were not

260

explicitly included because they were assumed identical in the homogeneously mixed sediments

261

embedded randomly at the same site. After correction for co-variation among explanatory

262

variables, the net variation explained by MWCNT treatments, time, and macrophytes was

263

substantial and highly significant (Table 1). Most importantly, MWCNT treatments explained

264

9.9% of the variation in the benthic community composition (p = 0.012). Exposure time

265

explained 35.0% of the variation (p = 0.002) and macrophyte abundance explained another 5.2%

266

(p = 0.004). The importance of time can be understood primarily from the fact that the

267

experiment was a recolonization experiment. At start there was no community at all, which

268

explains the importance of time in rebuilding the community. However, also small changes in

269

sediment composition (eg., RC) theoretically may have contributed to the significance of the

270

variability explained by time. The influence of macrophytes is well understood, because the

271

presence of macrophytes can change the composition of the sediment by rooting, by providing

272

refugee for invertebrates against predation or by reducing the incidence of light on the sediment.

273

Spatial distribution appeared not to be statistically significant (2.2% of the variation; p = 0.760),

274

which confirms that random positioning of the trays in the ditch did not influenced the

275

community composition caused by dissimilarities between trays. This includes the absence of

276

effects from small spatial dissimilarities in sediment characteristics (OC, RC), if any.

277

The first two axes of the variation in samples and species data revealed by RDA

278

explained 38.4% and 5.5%, respectively. The first axis shows a separation of the samples based

279

on exposure time. Samples retrieved after 3 months are on the left-hand side of the plot, whereas



Page 14 of 28 14

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

280

the trays exposed for 15 months are on the right-hand side of Figure 2a. The second axis relates

281

to the MWCNT level and macrophyte content. A clear separation between the controls (C0) in

282

the lower part of the figure and the contaminated samples (C1, C2 and C4) and the macrophyte

283

content in the upper part of the figure is visible. This difference between the controls and the

284

MWCNT dosed samples indicates an effect on the community structure.

285

Taxa that were located in the centre of the ordination plot (i.e. with short arrows) can be defined

286

as being hardly affected by MWCNT treatment or exposure time (Figure 2b). Taxa in the left

287

part of the diagram show a relatively strong relation with the 3 month samples. However, the

288

abundance of most of these taxa was low, so that the absence of these taxa after 15 months

289

should not be interpreted as a MWCNT treatment effect. The abundance of Sphaeriidae and

290

Hydracarina was high after 3 months, but these taxa had also a strong relation with the

291

macrophyte content. Dominant species, e.g. Lumbriculidae, Erpobdella, Sphaeriidae and

292

Chironomidae increased between 3 and 15 month exposure time despite MWCNT treatments,

293

which again suggests that time is the main factor in the recolonization process.

294

MWCNT treatments and taxa were not explicitly grouped or separated (Figure 2b). The higher

295

amount of macroinvertebrates after 15 months compared to 3 months of exposure suggests that

296

time and natural circumstances have a higher effect on the community composition than the

297

MWCNT sediment treatments. Nevertheless, the controls (C0) are clearly separated from the

298

rest, which confirm the influence of MWCNTs on the community structure.

299

Interestingly, the statistical analysis revealed a significant effect of a dose of up to 2 g/kg

300

MWCNTs on the invertebrate community. To our knowledge this is the first report of significant

301

community effects of carbon nanomaterials under long term field conditions. Although the

302

MWCNTs were added to the sediment in their pristine state, 15 months of aging in the sediment


Page 15 of 28


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

303

must have altered their aggregation state and possible transformation5, 13 such that they more

304

closely resemble an environmentally realistic state than those in current shorter term laboratory

305

tests. Current MWCNT characterization methodologies are insufficient to assess the exact

306

MWCNT state and concentration in sediments,13, 45 but the observed effects can be assumed to

307

relate to the aged and aggregated MWCNTs, rather than to the primary particles. However, the

308

exposure concentration of up to 2 g/kg is at least three orders of magnitude higher than exposure

309

concentrations that can be expected in natural sediments.5, 20 To test whether effects would still

310

be significant at lower MWCNT doses, we re-analysed the data four times, each time excluding

311

the highest dose (Table 1). The results show that despite the reduced statistical rigour due to the

312

smaller dataset, the effect of MWCNTs remains statistically significant, even when only the data

313

for the controls and the lowest treatment concentrations of 0.002 mg/kg are used. At this low

314

dose, MWCNTs still explained 7.1% of the community composition, a percentage that was

315

significant (p = 0.020). These data thus indicate that even the lowest MWCNT dose that was

316

tested (0.002 g/kg), seems to show significant chronic effects on the benthic community structure

317

after 15 months. We argue that this dose can be considered environmentally relevant because

318

model studies showed that given current emission rates, these concentrations can be reached

319

within decades,5, 20 or earlier on locations subject to emission from local point sources.

320

Although our study design could detect community effects and patterns of sensitivities of species

321

towards MWCNTs (Figure 2, discussed earlier), it was not fit to detect toxicity mechanisms for

322

individual species. One reason for this is the current lack of methods to characterize the in situ

323

state of MWCNTs in the sediment after 15 months.13, 45 Therefore MWCNT state, i.e. level of

324

aggregation, transformation and degradation, and concentration remained unknown and nominal

325

concentrations had to be used. Still, some general inferences can be made from the data and the



Page 16 of 28 16

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

326

literature. When RDA is performed using data from MWCNT treated systems only, i.e.

327

excluding the controls, the effect of MWCNT dose was not significant anymore (Table 1).

328

Furthermore, there was no clear trend in the percentage explained, between systems that differed

329

by orders of magnitude in MWCNT concentration (Table 1). This suggests that the nanoparticles

330

trigger an effect on (certain) species, which apparently is independent on dose. Dose-independent

331

bioaccumulation and effects of C60 have recently been reported for earthworms.46 Petersen et

332

al.13 reviewed MWCNT aquatic toxicity studies and reported minimal effects, i.e. effect ranges

333

of 0.3 – 300 g/kg for various benthic invertebrates and endpoints in 10 – 28 d sediment toxicity

334

tests. The observed toxicity was ascribed to MWCNT disturbances on epithelial surfaces,

335

oxidative stress, or blockage of the gastro intestinal tract. In our previous study, no effects of

336

0.002 – 2 g/kg MWCNTs on community composition were detected after 3 months.18 The main

337

difference of the present study with all these previous studies is that exposure time was much

338

longer. Long exposure and use of community composition as an endpoint for effect increase the

339

role of sub-lethal and behavioural effect mechanisms on the fitness of species. It has been

340

observed that certain benthic invertebrates sense and avoid contaminated sediment, which is a

341

common ecological mechanism.47 After 15 months the role of reproduction in the observed

342

community shifts must have been larger. We thus hypothesise that effects on reproduction of

343

invertebrates may have contributed to the observed community shifts. Whereas 51.7% of the

344

variation was explained by known variables, 48.3% of the variation remained unexplained.

345

Theoretically, this percentage can be explained by variables not explicitly included in this study,

346

like water (pH, conductivity, DOC, temperature, O2 content) and sediment quality, sediment

347

structure (OC and RC content, grain size), depth, habitat characteristics and traditional pollutants

348

(if any), but also biological factors like food quality and quantity, competition and predation.


Page 17 of 28


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

349

These factors are known to affect community composition,36 but were beyond the scope of this

350

work. Again, however, most of these variables were similar for all trays and therefore hardly will

351

have affected the detected trends.

352

MWCNT versus AC impacts on benthic community structure. The observed impacts of

353

MWCNTs on the benthic community structure can be compared directly to those of AC in our

354

recently published study, which was performed simultaneously in parallel systems in the

355

“Veenkampen”.19 Kupryianchyk et al19 reported an effect of 100 g/kg (i.e. 10%) AC of 9.0% (p =

356

0.008),19 which is very similar to the present effect observed for MWCNTs (9.9%, p = 0.012).

357

However, although these components have a comparable impact, the highest concentration of

358

MWCNTs was about fifty times lower than the highest AC concentration. This would imply that

359

the MWCNTs are about 50 times more potent in structuring the benthic community than AC. In

360

contrast, Kennedy et al22 reported a 10-20 times lower toxicity of high levels of MWCNTs

361

compared to AC. However, their test conditions, species and endpoints were very different from

362

those in the present study, rendering the data difficult to compare. Other recent studies propose

363

the use of AC as well as nanomaterials for in situ remediation.14, 48 Our findings suggest that

364

further research on the state, persistence, bioavailability and ecological safety of MWCNT

365

amendments to sediments is highly needed.

366

Implications. Whereas literature on short term laboratory tests shows relatively high effect

367

thresholds for MWCNTs,13 our present data suggest that they may be much lower in the long

368

term in the field. For MWCNTs, the observed effect threshold approaches concentrations

369

anticipated to occur in natural sediments. This illustrates the relevance of long term in situ effect

370

assessment of natural communities to nanomaterials. Although effect mechanisms are hard to

371

identify from field studies, the main benefit is that many naturally occurring factors that are not



Page 18 of 28 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

372

covered in short term laboratory tests are accounted for, and therefore automatically are

373

incorporated in the observed dose effect relationship. The present results also show that

374

multivariate endpoints may be more sensitive to community changes than univariate endpoints

375

such as the Shannon index, in which part of the community information is lost. Detection of in

376

situ effects of MWCNTs can have far reaching consequences for the risk assessment of

377

nanomaterials. We emphasize that although significant MWCNT effects were detected, the

378

current data primarily call for confirmation in follow-up field experiments.

379 380

Acknowledgements

381

We would like to thank Darya Kupryianchyk for her advice during the experiments and data

382

analysis. John Beijer and Erik Reichmann are acknowledged for their help and advice with

383

sediment sampling, placement and retrieving of the trays and species identification. We thank

384

Frits Gillissen for his help with the OC and RC measurements. Three anonymous reviewers are

385

acknowledged for their critical comments on the manuscript.

386 387

Supporting Information

388

Data on sediment contaminant concentrations, carbon characteristics, community measures and

389

taxa abundances of benthic macroinvertebrates. Additional figures on abundance of the benthic

390

community after 3 and 15 months. This information is available free of charge via the Internet at

391

http://pubs.acs.org/.

392


Page 19 of 28


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

393

References

394

(1)

395

Nowack, B.; von der Kammer, F., Paradigms to assess the environmental impact of

396

manufactured nanomaterials. Environ. Toxicol. Chem. 2012, 31, (1), 3-14.

397

(2)

398

Metcalfe, C.; Steevens, J. A.; Klaine, S. J.; Koelmans, A. A.; Horne, N., Ecotoxicity test methods

399

for engineered nanomaterials: Practical experiences and recommendations from the bench.

400

Environ. Toxicol. Chem. 2012, 31, (1), 15-31.

401

(3)

402

environment. Environ. Pollut. 2007, 150, (1), 5-22.

403

(4)

404

Mahendra, S.; McLaughlin, M. J.; Lead, J. R., Nanomaterials in the environment: Behavior, fate,

405

bioavailability, and effects. Environ. Toxicol. Chem. 2008, 27, (9), 1825-1851.

406

(5)

407

carbon nanoparticle concentrations in aquatic sediments. Environ. Pollut. 2009, 157, (4), 1110-

408

1116.

409

(6)

410

nanoparticle risk forecasting: Model formulation and baseline evaluation. Sci. Total. Envir. 2012,

411

426, (0), 436-445.

412

(7)

413

Heuvel-Greve, M. J.; Koelmans, A. A., Sediment toxicity testing of organic chemicals in the

414

context of prospective risk assessment: A review. Crit. Rev.Env. Sci. Technol. 2013,

415

DOI:10.1080/01496395.2012.718945.

Klaine, S. J.; Koelmans, A. A.; Horne, N.; Carley, S.; Handy, R. D.; Kapustka, L.;

Handy, R. D.; Cornelis, G.; Fernandes, T.; Tsyusko, O.; Decho, A.; Sabo-Attwood, T.;

Nowack, B.; Bucheli, T. D., Occurrence, behavior and effects of nanoparticles in the

Klaine, S. J.; Alvarez, P. J. J.; Batley, G. E.; Fernandes, T. F.; Handy, R. D.; Lyon, D. Y.;

Koelmans, A. A.; Nowack, B.; Wiesner, M. R., Comparison of manufactured and black

Money, E. S.; Reckhow, K. H.; Wiesner, M. R., The use of Bayesian networks for

Diepens, N. J.; Arts, G. H. P.; Brock, T. C. M.; Smidt, H.; Van den Brink, P. J.; Van den



Page 20 of 28 20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

416

(8)

417

nanomaterials in the aquatic environment. Anal. Bioanal. Chem. 2009, 393, (1), 81-95.

418

(9)

419

ecotoxicology and chemistry of manufactured nanoparticles. Ecotoxicology 2008, 17, (4), 287-

420

314.

421

(10)

422

tests of some nanomaterials. Environ. Toxicol. Chem. 2008, 27, (9), 1942-1947.

423

(11)

424

Horne, N.; Koelmans, A. A.; Klaine, S. J., Potential scenarios for nanomaterial release and

425

subsequent alteration in the environment. Environ. Toxicol. Chem. 2012, 31, (1), 50-59.

426

(12)

427

organisms respond to, and influence, the concentration of titanium dioxide nanoparticles? A

428

mesocosm study with algae and herbivores. Environ. Toxicol. Chem. 2012, 31, (10), 2414-2422.

429

(13)

430

Nguyen, T.; Huang, Q.; Henry, T. B.; Holbrook, R. D.; Chen, K. L., Potential Release Pathways,

431

Environmental Fate, And Ecological Risks of Carbon Nanotubes. Environ. Sci. Technol. 2011,

432

45, (23), 9837-9856.

433

(14)

434

the Benefits and Potential Risks. Environmental Health Perspectives 2009, 117, (12), 1823-1831.

435

(15)

436

Applications and emerging opportunities. Critical Reviews in Microbiology 2008, 34, (1), 43-69.

437

(16)

438

drinking water treatment processes. Water Research 2009, 43, (9), 2463-2470.

Farre, M.; Gajda-Schrantz, K.; Kantiani, L.; Barcelo, D., Ecotoxicity and analysis of

Handy, R. D.; von der Kammer, F.; Lead, J. R.; Hassellov, M.; Owen, R.; Crane, M., The

Velzeboer, I.; Hendriks, A. J.; Ragas, A. M. J.; Van de Meent, D., Aquatic ecotoxicity

Nowack, B.; Ranville, J. F.; Diamond, S.; Gallego-Urrea, J. A.; Metcalfe, C.; Rose, J.;

Kulacki, K. J.; Cardinale, B. J.; Keller, A. A.; Bier, R.; Dickson, H., How do stream

Petersen, E. J.; Zhang, L.; Mattison, N. T.; O’Carroll, D. M.; Whelton, A. J.; Uddin, N.;

Karn, B.; Kuiken, T.; Otto, M., Nanotechnology and in Situ Remediation: A Review of

Theron, J.; Walker, J. A.; Cloete, T. E., Nanotechnology and water treatment:

Hyung, H.; Kim, J. H., Dispersion of C-60 in natural water and removal by conventional


Page 21 of 28


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

439

(17)

440

Sedimentation of nanomaterials in natural waters. In SETAC World congress, Berlin, Germany,

441

May 2012; Vol. Abstract book p49.

442

(18)

443

effects of carbon nanotubes in aquatic sediments. Environment International 2011, 37, (6), 1126-

444

1130.

445

(19)

446

T. C.; Koelmans, A. A., Long-Term Recovery of Benthic Communities in Sediments Amended

447

with Activated Carbon. Environ. Sci. Technol. 2012, 46, (19), 10735-10742.

448

(20)

449

Concentrations of Engineered Nanomaterials (TiO2, ZnO, Ag, CNT, Fullerenes) for Different

450

Regions. Environ. Sci. Technol.2009, 43, (24), 9216-9222.

451

(21)

452

environment. Environ. Sci. Technol. 2008, 42, (12), 4447-4453.

453

(22)

454

J. C.; Weiss Jr., C. A., Factors influencing the partitioning and toxicity of nanotubes in the

455

aquatic environment. Environ. Toxicol. Chem. 2008, 1932–1941

456

(23)

457

expression in daphnids exposed to manufactured nanoparticles: Changes in toxicity with

458

functionalization. Environ. Pollut. 2009, 157, (4), 1152-1156.

459

(24)

460

Carbon Nanomaterials in Drosophila: Larval Dietary Uptake Is Benign, but Adult Exposure

461

Causes Locomotor Impairment and Mortality. Environ. Sci. Technol.2009, 43, (16), 6357-6363.

Quik, J. T. K., Velzeboer, I., Wouterse, M., Koelmans, A.A., van de Meent, D.,

Velzeboer, I.; Kupryianchyk, D.; Peeters, E. T. H. M.; Koelmans, A. A., Community

Kupryianchyk, D.; Peeters, E. T. H. M.; Rakowska, M. I.; Reichman, E. P.; Grotenhuis, J.

Gottschalk, F.; Sonderer, T.; Scholz, R. W.; Nowack, B., Modeled Environmental

Mueller, N. C.; Nowack, B., Exposure modeling of engineered nanoparticles in the

Kennedy, A. J.; Hull, M. S.; Steevens, J. A.; Dontsova, K. M.; Chappell, M. A.; Gunter,

Klaper, R.; Crago, J.; Barr, J.; Arndt, D.; Setyowati, K.; Chen, J., Toxicity biomarker

Liu, X. Y.; Vinson, D.; Abt, D.; Hurt, R. H.; Rand, D. M., Differential Toxicity of



Page 22 of 28 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

462

(25)

463

Ecotoxicology of carbon-based engineered nanoparticles: Effects of fullerene (C-60) on aquatic

464

organisms. Carbon 2006, 44, (6), 1112-1120.

465

(26)

466

nanomaterial suspensions to Daphnia magna. J. Nanopart. Res. 2009, 11, (1), 67-75.

467

(27)

468

B., Toxicity of carbon nanotubes to freshwater aquatic invertebrates. Environ. Toxicol. Chem.

469

2012, 31, (8), 1823-1830.

470

(28)

471

of the dilute sedimentary soot phase: Implications for PAH speciation and bioavailability.

472

Environ. Sci. Technol. 1997, 31, (1), 203-209.

473

(29)

474

polychlorinated biphenyls to soot and soot-like materials in the aqueous environment

475

mechanistic considerations. Environ. Sci. Technol. 2002, 36, (17), 3725-3734.

476

(30)

477

voor zoetwatermacroinvertebraten en methoden ter bepaling van de waterkwaliteit; Stichting

478

Leefmilieu: Antwerpen, Belgium,1991.

479

(31)

480

van Nederland; Koninklijke Nederlandse Natuurhistorische Vereniging: Utrecht, the Netherlands

481

1992.

482

(32)

483

Educatie); PIME: Lier, Vlaanderen, Belgium, 2003.

Oberdorster, E.; Zhu, S. Q.; Blickley, T. M.; McClellan-Green, P.; Haasch, M. L.,

Zhu, X. S.; Zhu, L.; Chen, Y. S.; Tian, S. Y., Acute toxicities of six manufactured

Mwangi, J. N.; Wang, N.; Ingersoll, C. G.; Hardesty, D. K.; Brunson, E. L.; Li, H.; Deng,

Gustafsson, O.; Haghseta, F.; Chan, C.; MacFarlane, J.; Gschwend, P. M., Quantification

Jonker, M. T. O.; Koelmans, A. A., Sorption of polycyclic aromatic hydrocarbons and

de Pauw N.; Vannevel, R. Macro-invertebraten en waterkwaliteit. Determineersleutels

Drost, M. B. P.; Cuppen, H.P.J.J.; van Nieukerken, E.J.; Schreijer, M. De waterkevers

Keustermans, K.; Quanten, E. Wateronderzoek PIME (Provinciaal Instituut voor Milieu


Page 23 of 28


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

484

(33)

485

U.K., 1988.

486

(34)

487

J., Evaluation of bioassays versus contaminant concentrations in explaining the

488

macroinvertebrate community structure in the Rhine-Meuse delta, the Netherlands. Environ.

489

Toxicol. Chem. 2001, 20, (12), 2883-2891.

490

(35)

491

in structuring in situ macroinvertebrate community composition along a salinity gradient.

492


493

(36)

494

structure in relation to food and environmental variables. Hydrobiologia 2004, 519, (1-3), 103-

495

115.

496

(37)

497

availability and effects on benthic community structure in floodplain lakes. Environ. Toxicol.

498

Chem. 2004, 23, (3), 668-681.

499

(38)

500

biological indicator of ecological and toxicological factors in Lake Saint-Francois (Quebec).

501

Environ. Pollut. 1996, 91, (1), 65-87.

502

(39)

503

Group: Wageningen, The Netherlands, 1990.

504

(40)

505

Windows user's guide: software for canonical community ordination (version 4.5); Ithaca

506

Microcomputer Power: New York, USA, 2002.

Magurran, A. E.; Ecological diversity and its measurement; University Press: Cambridge,

Peeters, E. T. H. M.; Dewitte, A.; Koelmans, A. A.; van der Velden, J. A.; den Besten, P.

Peeters, E. T. H. M.; Gardeniers, J. J. P.; Koelmans, A. A., Contribution of trace metals

Peeters, E. T. H. M.; Gylstra, R.; Vos, J. H., Benthic macroinvertebrate community

van Griethuysen, C.; van Baren, J.; Peeters, E. T. H. M.; Koelmans, A. A., Trace metal

PinelAlloul, B.; Methot, G.; Lapierre, L.; Willsie, A., Macroinvertebrate community as a

Ter Braak, C. J. F. Update notes: CANOCO version 3.1; Agricultural Mathematics

Ter Braak, C. J. F.; Smilauer, P. CANOCO reference manual adn CanoDraw for



Page 24 of 28 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

507

(41)

508

Gustafsson, O., Black carbon: The reverse of its dark side. Chemosphere 2006, 63, (3), 365-377.

509

(42)

510

nanotubes to chemothermal oxidation used to isolate soots from environmental samples.

511

Environ. Pollut. 2009, 157, (4), 1065-1071.

512

(43)

513

of the macrozoobenthos on large defaunated plots on a tidal flat in the Wadden Sea. J. Sea. Res.

514

1999, 42, (3), 235-254.

515

(44)

516

Cambridge University Press: Cambridge, U.K., 2003.

517

(45)

518

S. J.; Koelmans, A. A.; Horne, N.; Unrine, J. M., Analysis of engineered nanomaterials in

519

complex matrices (environment and biota): General considerations and conceptual case studies.

520


521

(46)

522

population. Ph.D Dissertation, Wageningen University, Wageningen, the Netherlands, 2013.

523

(47)

524

hydrocarbon-contaminated sediments by the freshwater invertebrates Gammarus pulex and

525

Asellus aquaticus. Environ. Toxicol. Chem. 2006, 25, (2), 452-457.

526

(48)

527

situ remediation of contaminated sediments using carbonaceous materials. Environ. Toxicol.

528

Chem. 2012, 31, (4), 693-704.

Koelmans, A. A.; Jonker, M. T. O.; Cornelissen, G.; Bucheli, T. D.; Van Noort, P. C. M.;

Sobek, A.; Bucheli, T. D., Testing the resistance of single- and multi-walled carbon

Beukema, J. J.; Flach, E. C.; Dekker, R.; Starink, M., A long-term study of the recovery

Leps, J., Smilauer, P., Multivariate Analysis of Ecological Data Using CANOCO;

von der Kammer, F.; Ferguson, P. L.; Holden, P. A.; Masion, A.; Rogers, K. R.; Klaine,

van der Ploeg, M. Unravelling hazards of nanoparticles to earthworms, from gene to

De Lange, H. J.; Sperber, V.; Peeters, E. T. H. M., Avoidance of polycyclic aromatic

Rakowska, M. I.; Kupryianchyk, D.; Harmsen, J.; Grotenhuis, T.; Koelmans, A. A., In

529


Page 25 of 28


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

530

Figures and Tables

531

Table 1. Explained variation of community structure for MWCNT addition, macrophyte

532

abundance and spatial distribution with significance level for redundancy analysis (RDA).

533

MWCNT doses Explanatory Variation explained by Significance of explained considered Variables explanatory variable (%) variation (p value)a 0 – 2 g/kg MWCNTs 9.9 0.012* macrophytes 5.2 0.004** spatial distr. 2.2 0.760 time 35.0 0.002** unexplained 48.3 0 – 0.2 g/kg MWCNTs 10.6 0.002** macrophytes 6.2 0.004** spatial distr. 2.8 0.840 time 36.7 0.002** unexplained 45.8 0 – 0.02 g/kg MWCNTs 17.9 0.002** macrophytes 4.9 0.022* spatial distr. 3.4 0.726 time 39.2 0.002** unexplained 38.4 0 – 0.002 g/kg MWCNTs 7.1 0.020* macrophytes 6.2 0.056 spatial distr. 7.3 0.244 time 44.3 0.002** unexplained 36.1 0.002 – 2 g/kg MWCNTs 6.7 0.392 macrophytes 4.1 0.054 spatial distr. 4.7 0.350 time 28.8 0.002** unexplained 51.0 a statistical significance was tested using Monte Carlo permutation test with 499 permutations.

534

*Significant at p < 0.05

535

**Significant at p < 0.01

536



Page 26 of 28 26

537

a

16 3m

14

15m

OC (%)

12

10 8 6 4 2 0 0

0.002 0.02 0.2 nominal concentration MWCNT (g/kg)

2

538

b

0.8 3m 0.7

15m

0.6 0.5 RC (%)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

0.4 0.3 0.2 0.1 0 0

0.002 0.02 0.2 nominal concentration MWCNT (g/kg)

2

539 540

Figure 1. Organic carbon (OC) (a) and residual carbon (RC) (b), measured with chemothermal

541

oxidation (CTO375) in sediment samples after 3 months (white bars) and 15 months (black bars)

542

of exposure. SD’s (n=3) do not relate to variability among replicates in the field, but to

543

analytical variability among repeated measures on mixed samples per treatment.

544 545


Page 27 of 28


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

a

b

546 547

Figure 2. Samples scores (a) and macroinvertebrate scores (b) of the first two axes of a

548

redundancy analysis (RDA) of the benthic community from the recolonization study performed

549

in the Veenkampen with sediment contaminated with 0 (circles, C0), 0.002 (squares, C1), 0.02

550

(diamonds, C2), 0.2 (triangles, C3) and 2 (stars, C4) g/kg MWCNTs after 3 (open symbols) and

551

15 months (grey symbols). Environmental data are indicated with triangles. Samples and taxa

552

grouped in the same part of the biplot are related to one another. The lengths of the taxa arrows

553

indicate the strength of the relationship.

554 555 556



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

TOC/Abstract Art


Page 28 of 28

Multiwalled Carbon Nanotubes at Environmentally Relevant

Recommend Documents