Ocean acidification increases the toxicity of contaminated sediments

Ocean acidification increases the toxicity of contaminatedsedimentsDAV ID A . ROBERTS * † 1 , S I LVANA N . R . B I RCHENOUGH* 1 , CER I LEW I S ‡ 1 ,

MATTHEW B . SANDERS § , TH I BOLAM* and DAVE SHEAHAN*

*Centre for Environment, Fisheries & Aquaculture Science, Lowestoft, NR33 0HT, UK, †School of Marine and Tropical Biology,

James Cook University, Townsville, 4811, Australia, ‡College of Life & Environmental Sciences, University of Exeter, Exeter,

EX4 4QD, UK, §Centre for Environment, Fisheries & Aquaculture Science, Weymouth, DT4 8UB, UK

Abstract

Ocean acidification (OA) may alter the behaviour of sediment-bound metals, modifying their bioavailability and

thus toxicity. We provide the first experimental test of this hypothesis with the amphipod Corophium volutator.

Amphipods were exposed to two test sediments, one with relatively high metals concentrations (Σmetals

239 mg kg�1) and a reference sediment with lower contamination (Σmetals 82 mg kg�1) under conditions that mimic

current and projected conditions of OA (390–1140 latm pCO2). Survival and DNA damage was measured in the

amphipods, whereas the flux of labile metals was measured in the sediment and water column (WC) using Diffu-

sive Gradients in Thin-films. The contaminated sediments became more acutely toxic to C. volutator under elevated

pCO2 (1140 latm). There was also a 2.7-fold increase in DNA damage in amphipods exposed to the contaminated

sediment at 750 latm pCO2, as well as increased DNA damage in organisms exposed to the reference sediment, but

only at 1140 latm pCO2. The projected pCO2 concentrations increased the flux of nickel and zinc to labile states in

the WC and pore water. However, the increase in metal flux at elevated pCO2 was equal between the reference and

contaminated sediments or, occasionally, greater from reference sediments. Hence, the toxicological interaction

between OA and contaminants could not be explained by effects of pH on metal speciation. We propose that the

additive physiological effects of OA and contaminants will be more important than changes in metal speciation in

determining the responses of benthos to contaminated sediments under OA. Our data demonstrate clear potential

for near-future OA to increase the susceptibility of benthic ecosystems to contaminants. Environmental policy

should consider contaminants within the context of changing environmental conditions. Specifically, sediment

metals guidelines may need to be reevaluated to afford appropriate environmental protection under future

conditions of OA.

Keywords: contaminated sediment, Corophium volutator, DNA damage, metals, ocean acidification, toxicity

Received 2 August 2012; revised version received 25 September 2012 and accepted 25 September 2012

Introduction

Marine habitats are changing at an unprecedented

rate in terms of sea surface temperature, sea ice

cover, salinity, alkalinity, pH and ocean circulation

(Bulling et al., 2010; Rogers & Laffoley, 2011; Honisch

et al., 2012). Increased emissions of carbon dioxide

(CO2) as a result of anthropogenic activities are pre-

dicted to cause rising atmospheric and oceanic tem-

peratures with direct implications for the ecology of

terrestrial and marine ecosystems (Sabine et al., 2004;

Turley et al., 2010). The world’s oceans are a major

sink for anthropogenic CO2 and the effects of

increased CO2 dissolution in seawater on fundamen-

tal acid-base equilibria are well understood (Gattuso

& Hansson, 2011). However, CO2-induced changes in

physicochemical attributes of the oceans are not

occurring in isolation. Anthropogenic pollution, for

example, is a continuing threat to the marine environ-

ment (Rogers & Laffoley, 2011). It has become appar-

ent that climate change and CO2 emissions may have

indirect effects on marine ecosystems, and may also

interact with concurrent stressors or other natural

phenomena. Climate change may increase the suscep-

tibility of marine species to disease and marine com-

munities to invasion by exotic species, or alter the

bioavailability and toxicity of contaminants. This lat-

ter indirect effect of climate change has been the

focus of much recent research, with several excellent

syntheses available (Macdonald et al., 2005; Noyes

et al., 2009). Work in this field has focused largely on

the influence of temperature on the behaviour of

Correspondence: Dave Sheahan, tel. + 44 01502 524 535,

fax + 44 01502 513 865, e-mail: [email protected]; Silvana

Birchenough, tel. + 44 01502 527786, fax + 44 01502 513 865,

e-mail: [email protected] authors contributed equally to the preparation of this

manuscript.

340 © 2012 Blackwell Publishing Ltd

Global Change Biology (2013) 19, 340–351, doi: 10.1111/gcb.12048

contaminants in marine systems and the cycling of

contaminants in the atmosphere. Comparatively, little

research has considered how ocean acidification (OA)

may influence the fate and effects of contaminants in

marine ecosystems.

Metals are one of the most common types of coastal

contaminant and are found in high concentrations in

the waters and sediments of many coastal and estuarine

systems (Bryan & Langston, 1992). OA is expected to

alter the bioavailability of water-borne metals (Millero

et al., 2009). The toxic free-ion concentration of metals

such as copper (Cu) may increase by as much as 115%

in coastal waters in the next 100 years due to reduced

pH (Pascal et al., 2010; Richards et al., 2011), whereas

the free-ion concentration of other metals including

cadmium (Cd) may decrease or be unaffected (Lacoue-

Labarthe et al., 2009, 2011, 2012; Pascal et al., 2010). One

might therefore predict greater metal toxicity in organ-

isms with exposure under higher pCO2. This hypothe-

sis is supported by the observed influence of increased

pCO2 on the bioaccumulation of trace metals in the eggs

and embryos of the squid Loligo vulgari (Lacoue-

Labarthe et al., 2011, 2012) and eggs of the cuttlefish

Sepia officinalis (Lacoue-Labarthe et al., 2009). While

these studies have measured bioaccumulation, the only

current study to investigate the influence of OA on

metal toxicities showed increased toxicity of Cu, but

not Cd, to the copepod Amphiascoides atopus under con-

ditions of elevated pCO2 (Pascal et al., 2010). This early

study is suggestive of interactions between aqueous

metals and OA. Even less is currently known about

how OA may influence the behaviours of metals bound

to sediments (Royal Society, 2005; Millero et al., 2009).

A recent report examined the role of elevated CO2 in

controlling fluxes of labile metals from contaminated

sediments (Ardelan et al., 2009). However, this experi-

ment focused on the failure of CO2 storage caverns and

therefore considered pCO2 concentrations far in excess

of those predicted to occur through OA from atmo-

spheric carbon sources. The implications of such altered

metal fluxes for the health of sediment-dwelling biota

have not been considered to date.

Metals such as Cu and zinc (Zn) are essential for a

number of biochemical processes, but are toxic at

elevated concentrations mostly via the production of

reactive oxygen species (ROS) (Stohs & Bagchi, 1995;

Valko et al., 2005) or covalent binding of the metal ion

to macromolecules. DNA damage associated with ROS

is thought to be a major source of genomic instability in

a living cell (Doudican et al., 2005). DNA damage there-

fore provides a sensitive endpoint for toxicological

studies of metal contamination (Jha, 2008), and has

been used in numerous case studies as a sensitive mea-

sure of sublethal effects of contaminated sediments on

the health of marine invertebrate infauna, including

amphipods (Neuparth et al., 2005) and polychaetes

(Lewis & Galloway, 2008).

The biological responses of organisms to OA are

relatively difficult to predict and experimental studies

have produced a range of results. Until recently, OA

research mainly considered the effect of declining pH

and perturbed carbonate chemistry on a range of

physiological processes, initially focusing on the calcifi-

cation process of marine invertebrates such as crusta-

ceans, echinoderms, coccolithophores and corals. For

example, shell formation in a variety of calcifying

species may be affected by OA and this could have

global repercussions for carbonate production in the

ocean (Lebrato et al., 2010). More recently, other pH-

dependent processes such as acid-base regulation,

metabolism and reproduction have also been studied

(e.g. Spicer et al., 2011; Stumpp et al., 2012). Some

meta-analyses have shown that taxa vary in their sensi-

tivity to OA, and that some organisms may be more

resilient than commonly predicted due to compensa-

tory biological processes and small-scale spatial and

temporal variability in ocean pH (Hendriks et al.,

2010). There are, however, discrepancies with these

approaches suggesting that meta-analyses could be

masking subtle variations in response, such as sensitiv-

ity to OA in different life stages (Dupont et al., 2010).

More recently attention has shifted to the potential

impact of OA on energetic partitioning between differ-

ent physiological processes (Wood et al., 2008; Pascal

et al., 2010). Increasing evidence suggests that exposure

to OA can lead to metabolic depression (Miles et al.,

2007), reduced growth (Michaelidis et al., 2005) and

reduced energy reserves (Langenbuch & Portner, 2002,

2003), particularly in species with poor ion regulation

(Whiteley, 2011). Thus, even species that show no

direct effects of OA on calcification may exhibit physio-

logical and behavioural trade-offs (Findlay et al., 2011).

It is therefore important to measure sensitive responses

in organisms to understand the nature of OA effects,

particularly when investigating responses to simulta-

neous stressors such as OA and contaminants.

This study had two primary aims. First, we charac-

terized the influence of near-future OA on fluxes of

labile trace metals from two field-collected sediments

with different physical properties and trace metal con-

centrations. Fluxes were measured from sediments to

the pore water (PW), sediment–water interface (SWI)

and overlying water (OW) under both current and

projected pCO2 scenarios. Second, we examined the

chronic and acute toxicity of contaminated sediments

under a range of pCO2 concentrations using the stan-

dard sediment toxicity test species Corophium volutator,

to determine the potential interactions between OA

© 2012 Blackwell Publishing Ltd, Global Change Biology, 19, 340–351

OCEAN ACIDIFICATION AND CONTAMINATED SEDIMENT 341

and contaminant exposure in toxicological responses

of benthic infauna.

Materials and methods

Study site

The test sediments were of two types; relatively uncontami-

nated reference sediment and nominally contaminated sedi-

ment (Table 1). The reference sediment was collected from

Shoebury Sands, near Southend-on-Sea, Essex, UK (51°31′49″N, 0°48′14″W). Contaminated sediments were collected from

the West Inner Tees dredged material disposal location, Uni-

ted Kingdom (IT4; 54°40′80″N, 1°03′00″W). The Centre for

Environment, Fisheries and Aquaculture Science (Cefas) has a

statutory duty to undertake annual monitoring of dredged

material disposal in this area. The Tees Estuary has been

subject to significant inputs of metals from mining activity

(Plater et al., 1999) and sewage effluent and industrial waste

from various chemical industries (Hardy et al., 1993). The most

significant inputs occurred during the 1970s, but recent studies

have also reported elevated concentrations of metals, polycy-

clic aromatic hydrocarbons (PAHs) and brominated flame

retardants in sediments and biota (Morris et al., 2004; Bolam

et al., 2011). The eroding mud flat within the estuary is also a

source of contamination due to high concentrations of lead

(Pb) and zinc (Zn) in the intertidal sediments.

Test sediment and organism collection

The test sediments were collected in May 2011 from a research

vessel using a 0.10 m2 day grab. Sediments were scooped

from the upper 5 cm of the grab with an acid-washed plastic

spoon and stored in clean plastic buckets. Sediment was not

collected that was in contact with the grab surface. Sediments

were then sieved through a 500 lm sieve and stored at 10 °Con the research vessel for approximately 24 h, then blast fro-

zen and stored at �20 °C until use in the experiments (Febru-

ary 2012). The sediments were defrosted prior to the test by

slowly bringing them to 4 °C over a period of 3 days in the

buckets in which they were collected. Amphipods were col-

lected from an intertidal mudflat at Dalgety Bay, Fife (56°02′19″N, 3°20′03″W). The top 5 cm of sediment was scooped with

a hand spade and gently sieved through a 500 lm mesh in situ

to retain adults but remove neonates (Roddie & Thain, 2001).

The amphipods were then placed in a plastic bag with detrital

material and site water and shipped overnight to the labora-

tory in a cool box. On arrival at the laboratory, the amphipods

were gradually acclimated to test temperature (15 ± 1 °C) andsalinity (30 ± 1 ppt) over a period of 5 days. Mortality during

this time was less than 10%.

Experimental facilities

Experiments were conducted in the Ocean Acidification

Experimental Facility at Cefas in Weymouth, United Kingdom

(UK). Manipulation of pCO2 was provided by a compressed

air supply passed through a dew point drying and filtration

system. The pCO2 content was measured using a calibrated

analyser and CO2 (certified 99.5% food grade, BOC) was

added to achieve the desired pCO2. Nominal gas mixtures

were 390, 750 and 1140 latm pCO2. These gas mixtures were

adopted to mimic current and projected future seawater car-

bonate chemistry as stated in EU ocean acidification research

recommendations (Riebesell et al., 2010). Seawater was

pumped from a coastal inlet pipe through a 0.45 lm UV steril-

iser and into 4 columns (0.2 m diameter 9 2.2 m height). Gas

flow was added at the base of each column via a ceramic fine

bubble diffuser. An additional bleed valve permitted indepen-

dent measurement of gas mixtures using a CO2/H2O gas ana-

lyser calibrated against certified CO2 gas mixtures (0 and

2000 ppm, BOC, UK).

Experimental procedures

Toxicity tests were undertaken to assess the interactive effects

of reduced pH (due to elevated pCO2) and contaminated sed-

iments on survival and DNA damage in C. volutator. Experi-

mental procedures followed existing protocols for the

10 days C. volutator toxicity test (Roddie & Thain, 2001), with

adaptations for the purposes of this experiment. Specifically,

the exposures were reduced to 9 days, and were conducted

in 20 cm long clear Perspex cores (8 cm diameter, Fig. 1).

Two days prior to the experiments, the sediments were

homogenized with an acid-washed plastic hand shovel and

added to the cores to a depth of 12 cm. The cores were over-

laid with water in one of 12 glass aquaria and left to equili-

brate for 48 h. There were six cores for each level of the

‘sediment 9 pCO2’ interaction, and these were split evenly

between two 40-l glass aquaria per treatment. Clean and con-

taminated sediments were not housed in the same aquaria to

avoid cross-contamination.

Table 1 Trace metal and metalloid concentrations, particle

size and organic carbon contents in the reference and contami-

nated test sediments. Metals and OCN data are mean concen-

trations (mg kg�1 dry weight) ± SE (n = 3 samples per

sediment type). Bold values exceed the applicable effects

range low (ERL) for that metal (Buchman, 2008)

Shoebury sands

(‘Reference’)

Inner tees

(‘Contaminated’)

As 5.33 ± 0.3 12.67 ± 0.3

Cd 0.16 ± 0.01 0.23 ± 0.01

Cr 8.83 ± 0.8 40.33 ± 1.2

Cu 7.93 ± 1.0 26.67 ± 1.9

Fe 7958 ± 631 24976 ± 1938

Mn 115.00 ± 8.6 433.00 ± 30.9

Ni 5.37 ± 0.4 27.00 ± 3.2

Pb 24.00 ± 0.6 44.00 ± 1.0

Zn 30.33 ± 2.9 88.33 ± 2.6

% Sand 94.24 ± 0.1 51.78 ± 1.4

% Fines 5.76 ± 0.1 47.85 ± 1.2

OCN 0.2 ± 0.01 3.8 ± 0.2

© 2012 Blackwell Publishing Ltd, Global Change Biology, 19, 340–351

342 D. A. ROBERTS et al.

Diffusive Gradients in Thin-films (DGTs) were used to

monitor labile metals in the overlying water and throughout

the sediment profiles during the experiments. The DGT units

were deployed in the cores on the first day of the experiment.

Half of the cores received a DGT probe whereas the remaining

replicates received a DGT disk which was attached to the side

of the Perspex core approximately 5 cm above the SWI. A

height of 5 cm was chosen as C. volutator is a bioturbating spe-

cies that is capable of resuspending fine particulates in a shal-

low boundary layer above the SWI (de Deckere et al., 2000; de

Backer et al., 2011). Holes were drilled into the cores just above

to SWI and covered with 300 lm mesh to encourage water

exchange across the sediment surface and prevent the test ani-

mals from escaping (Fig. 1).

Individual amphipods 5–7 mm long were randomly

assigned to each core (20 amphipods per core). At the end of

the experiment, the number of surviving amphipods was

counted for each replicate. The DGT samplers were processed

approximately 1 week after the experiment ended. For the

sediment probes, the resin gel layer was sliced at 1 cm inter-

vals using a razor blade. Each individual DGT sample was

placed in 1 ml of 1% HNO3 and left to elute for 24 h.

COMET assay

Two live amphipods from each replicate were pooled in a

microcentrifuge tube containing 200 ll PBS buffer and homog-

enized using a microcentrifuge tube pestle and grinder. The

resulting homogenate was centrifuged to separate the larger

tissue debris from the cell suspension. The COMET assay was

performed using 20 ll of this cell suspension according to the

methods adapted by Lewis et al. (2008), using alkaline condi-

tions at 5 °C. Briefly, 1 h lysis followed by 45 min denatur-

ation in electrophoresis buffer (0.3 M NaOH and 1 mM EDTA,

at pH 13) and then electrophoresis for 30 min at 25 V and

300 mA followed by neutralization. Cells were stained with

20 mg L�1 ethidium bromide and examined using a fluores-

cent microscope with 420–490 nm excitation filter and a

520 nm emission filter. One hundred cells per preparation

were quantified using COMET Assay IV (Perceptive Instru-

ments®, Bury St Edmonds, UK).

Water quality and chemical analyses

Temperature and pH ([H+]) were logged every 30 min during

the exposure in one replicate tank from each treatment. Water

samples were taken from each tank on days 0, 5 and 9 and

passed through a 0.45 lm filter. The salinity and temperature

of each sample was recorded. Duplicate water samples were

collected and preserved for analysis of dissolved inorganic

carbon (DIC), total alkalinity (TA), total phosphate (TP) and

silicate (TSi) analysis. DIC and TA analyses were conducted at

the National Oceanographic Centre, Southampton, United

Kingdom, by colorimetric and closed-cell titration methodolo-

gies, respectively, according to Dickson et al. (2007). TP and

TSi were analysed at Cefas Lowestoft laboratory according to

Kirkwood (1996). Absolute seawater pH and pCO2 was calcu-

lated from TA and DIC with CO2SYSver. 14 using published

dissociation constants (Mehrbach et al., 1973; Dickson & Mille-

ro, 1987; Dickson, 1990; Lewis & Wallace, 1998). In addition,

water samples from the overlying water column (WC) were

taken at the conclusion of the experiment for analysis of

organics (PCB and PAH suite) and ammonia. For the DGT

samplers, trace metals (Cd, Cu, Fe, Mn, Ni, Pb and Zn) were

quantified by ICP-MS or ICP-AES following a 10- or 20-fold

dilution respectively. It is not possible to measure arsenic (As)

flux using the DGT samplers deployed in our experiments.

Sediment samples were digested in HNO3 using enclosed-ves-

sel microwave. The digests were diluted and analysis of trace

elements was performed by ICP-MS and ICP-AES.

Quality assurance and quality control

Trace metals were determined by ICP-MS or ICP-AES with

external calibration. Certified reference materials (CRM

PACS-2, CRM NWTH-2), standards and blank reagents were

run within each sample batch for quality control. Internal

Quality Control is based on Shewhart charts which are built

with results obtained from reference materials using the North

West Analytical Quality AnalystTM (Northwest Analytical

Inc., Portland, OR, USA). Warning and control limits are

Fig. 1 Positioning of overlying water (OW), sediment–water

interface (SWI) and pore water (PW) DGTs in experimental

cores. Sediment is depicted in dark grey.

© 2012 Blackwell Publishing Ltd, Global Change Biology, 19, 340–351

OCEAN ACIDIFICATION AND CONTAMINATED SEDIMENT 343

defined as 2r and 3r respectively. If the results obtained for

the CRMs are within the defined limits, the results for the

sample batch are accepted, if not they are rejected and the

batch reanalysed. In addition, blank samples were analysed to

determine potential contamination via the DGT samplers and

reagents at each stage. All survival counts and data entry for

the toxicological study were confirmed by two independent

operators. The COMET slides were scored within a week of

preparation and kept on ice between sampling to prevent any

further induction of DNA damage. All slides were scored

blind to avoid any potential operator bias.

Statistical analyses

Survival data were analysed by partly nested three-factor

Analysis of variance (ANOVA), the two fixed orthogonal factors

being ‘sediment’ (two levels; clean and contaminated) and

‘pCO2’ (three levels; 390, 750, and 1140 latm). In addition, the

factor ‘tank’ was nested within the interaction ‘sedi-

ment 9 pCO2’ to test for potential tank effects. As this factor

was non-significant for both survival and DNA damage in

C. volutator (P > 0.250), it was eliminated from the model and

the data were pooled to increase power for the test of main

effects and their interaction (Underwood, 1996). There were

insufficient numbers of surviving amphipods in one of the

experimental treatments to allow samples to be collected for

DNA damage (contaminated sediment 9 1140 ppm pCO2).

One replicate core from each tank was randomly assigned

either a sediment probe or DGT disk which negated the need

for the nested ‘tank’ term in the model. Metal flux data for the

overlying water (OW; 5 cm above the sediment surface) and

the sediment–water interface (SWI; 0.5 cm above the sediment

surface) were analysed using a two-factor ANOVA (‘sediment’

and ‘pCO2’, both fixed). Metal data from the DGT probes were

analysed using a three-factor ANOVA, the additional factor

being ‘depth’. When concentrations were less than the limit of

detection (LOD) for the relevant analytical procedure, it was

assumed that the value was half the LOD when plotting the

data. However, formal analyses were limited to samples for

which metal concentrations exceeded the LOD. In all cases,

normality and homogeneity of variance were tested by exam-

ining residual histograms and scatter plots of estimates vs.

residuals respectively (Quinn & Keough, 2002). When neces-

sary data were log transformed to satisfy the assumptions of

ANOVA.

Results

Fluxes of trace metals

Overlying water column and sediment–water interface. The

flux of dissolved nickel (Ni) increased to the OW and at

the SWI at a pCO2 of 1140 latm compared with 390 and

750 latm and there was no difference in Ni flux

between the reference and contaminant sediment

(Figs 2a and 3a respectively; S1). Increased pCO2 also

affected manganese (Mn) and iron (Fe) fluxes to OW

and SWI, respectively, but the response differed

between the reference and contaminated sediment as

shown by the significant ‘sediment 9 pCO2’ interaction

(S1). Fluxes of Mn to OW did not differ among pCO2

treatments in the reference sediment. However, fluxes

were enhanced from the contaminated sediment at

750 latm compared with 390 latm, with intermediate

fluxes at the highest pCO2 treatment (Fig. 2b). There

was no effect of sediments or pCO2 on dissolved Mn

fluxes to the SWI (Fig. 3c; S1). There was a significantly

greater flux of Fe to the SWI from the reference sedi-

ment under a pCO2 of 1140 latm.

Fluxes of lead (Pb), Cd and Cu differed between ref-

erence and contaminated sediments, and were unaf-

fected by pCO2. Fluxes of Cd and Cu to the OW were

greater from the reference sediments than from the con-

taminated sediments (Fig. 2d and e; S1). However, at

the SWI fluxes from the contaminated sediments were

greater for Cd, Cu and Zn (Fig. 3d, e and g respectively;

S1). In the OW, fluxes of dissolved Pb were significantly

greater from contaminated sediments, while at the SWI

the opposite was observed (Figs 2c and 3c respectively;

S1). Fluxes of Fe and Zn to OW were unaffected by the

pCO2 treatments and did not differ between the test

sediments (Fig. 2f and g respectively; S1). Chromium

(Cr) was below limits of detection in all OW, SWI and

pore water samplers.

Interstitial pore water. Flux of labile Ni to interstitial PW

was significantly higher in the 750 and 1140 latm treat-

ments than they were in the 390 latm treatment

(Fig. 4a; S2). The Mn flux to PW showed a similar pat-

tern to those measured in the OW, with a significantly

greater flux of Mn to PW at 750 latm than 390 latm.

The fluxes at 1140 latm did not differ from the 390

treatment (Fig. 4b; S2). Fe was also affected by pCO2

with higher flux from the sediment at 1140 than

750 latm (Figure S5b, S2). Mn flux to PW was also

significantly greater from the reference sediment than

the contaminated sediment, and decreased with sedi-

ment depth (Fig. 4b; S2).

Significant interactions between the sediment type,

pCO2 and depth were detected for Cd, Cu and Zn

flux to the PW (S2). Fluxes of dissolved Zn were

significantly higher from the reference sediments when

incubated at 750 and 1140 latm than at 390 latm.

There was, however, no significant difference in Zn flux

among the pCO2 treatments in the contaminated

sediment cores (Fig. 4c; S3). Fluxes of Cd, Cu and Zn

were all higher from the reference sediment than the

contaminated sediment in the upper 1–6 cm of the sedi-

ment cores, but at lower depths fluxes of metals were

the same from the two test sediments (Fig. 4d–e; S5aand, S4).

© 2012 Blackwell Publishing Ltd, Global Change Biology, 19, 340–351

344 D. A. ROBERTS et al.

Toxicological responses

The survival of C. volutator was not affected by pCO2 in

the reference sediment; however, in the contaminated

sediment, their survival was significantly lower at

a pCO2 of 1140 latm than at 390 and 750 latm (‘sedi-

ment 9 pCO2’; F2,30 = 3.374, P = 0.048, Fig. 5a).

Survival in the contaminated sediment at 1140 latmaveraged 6%, while the other two contaminated sedi-

ment treatments attained mean survival of 50–55%(Fig. 5a).

Due to insufficient survival of the amphipods, no

DNA damage data were obtained from one of the

treatments. An ANOVA was therefore conducted on the

data for the 390 and 750 latm treatments only. There

was a significant interaction between sediment and

pCO2 in this reduced dataset (‘sediment 9 pCO2’;

F1,20 = 5.67, P = 0.027). DNA damage did not differ

between the 390 and 750 latm treatments in the refer-

ence sediment. However, there was a significant

increase in DNA damage in C. volutator exposed to the

contaminated sediments at 750 latm relative to the con-

taminated sediments at 390 ppm (Fig. 5b). There was

also an increase in DNA damage in C. volutator

exposed to the reference sediments under the highest

pCO2 relative to both the 390 and 750 latm treatments

(Fig. 5b, one-tailed unpaired t-test P = 0.024).

pCO2 characterization

Measured pCO2 in the experimental treatments tended

to decline from the first to last day of the experiment

(Actual pCO2; ‘Time’, F2,18 = 30.52, P < 0.001, Table 2).

Mean pCO2 concentrations at the beginning of the exper-

iments were approximately 470, 750 and 1150 latm in

the nominal 390, 750 and 1140 treatments respectively.

There was, however, no significant difference between

pCO2 in the tanks containing reference and contami-

nated sediments. By the end of the experiment, mean

pCO2 in the tanks were 410, 702, and 943 latm (Table 2).

Again, there was no difference in mean pCO2 in the

reference and contaminated sediments.

(b)(a)

(d)(c)

(f)(e)

(g)

Fig. 2 Flux of labile metals from test sediments to the overlying water column. Data are mean metal fluxes (nmol

cm�1 sec�1) ± SE (n = 2). For significant ‘sediment 9 pCO2’ interactions, Tukey’s post hoc comparisons contrasted fluxes among differ-

ent levels of pCO2 within sediment types. Bars marked with common letters are not significantly different (P > 0.05).

© 2012 Blackwell Publishing Ltd, Global Change Biology, 19, 340–351

OCEAN ACIDIFICATION AND CONTAMINATED SEDIMENT 345

Quality assurance and quality control

Physicochemical measurements during the toxicological

experiment were within limits specified within the test

protocol (Roddie & Thain, 2001). Specifically, the experi-

ment-wide mean temperature was 15.9 °C and all point

measurements deviated by <2 °C from the target test

temperature of 15 °C. Dissolved oxygen averaged

8.6 mg l�1 (84–92% saturation). Mean salinity was

30.6 ppt and all point measurements were within 2 ppt

of the target test salinity of 30 ppt. Less than 10%mortal-

ity of the test organism occurred during the acclimation

period. All CRM and standards analyses were within

control chart limits for the chemical analyses, and metal

concentrations in the reagent blanks were below limits

of detection (LOD). Pb, Zn, Cu and Ni were detected in

some, but not all, of the DGT blanks at concentrations

slightly above the relevant LOD. In all cases, concentra-

tions were an order of magnitude lower than those

measured in the treatment DGTs from both reference

and contaminated sediments (Fig. 4). Cd, Cr, Fe and Mn

were all below LOD in all DGT blank samples.

Discussion

Simultaneous exposure to OA resulted in a greater

toxicity of contaminated sediment to C. volutator. This

was true for both the sublethal endpoint (DNA dam-

age) and acute toxicity. There was no effect of pCO2 on

the survival of C. volutator maintained in clean sedi-

ments, indicating that the OA stress induced in this

experiment was not acutely toxic. Amphipods were

more susceptible to acute toxicity at the highest pCO2

of 1140 latm than at 390 and 750 latm in the contami-

nated sediments. There was also a significant increase

in DNA damage at the intermediate pCO2 concentra-

tion (750 latm) in the contaminated sediments, while

there was no effect in the corresponding reference sedi-

ment at that pCO2. DNA damage is likely to have been

even greater at the highest pCO2 level in the contami-

nated sediments as amphipod survival was only 6%

(which precluded analysis). The highest pCO2 level

tested (1140 latm) was sufficient to increase the DNA

damage in C. volutator in the reference sediment, indi-

cating the OA treatment itself did have measurable

(b)(a)

(d)(c)

(f)(e)

(g)

Fig. 3 Flux of labile metals from test sediments to the sediment–water interface. Data are mean metal fluxes (nmol

cm�1 sec�1) ± SE (n = 2). Format as per Fig. 2.

© 2012 Blackwell Publishing Ltd, Global Change Biology, 19, 340–351

346 D. A. ROBERTS et al.

(b)(a)

(d)(c)

Fig. 4 Depth profiles of labile Ni, Mn, Zn and Cu flux from reference (grey lines) and contaminated sediments (black lines) to the pore

water. Data are mean metal fluxes (nmol cm�1 sec�1) ± SE (n = 2). Mean metal flux measured in the DGT blanks are depicted as a

cross on the x-axis for comparison. Results of post hoc comparisons and figures for Cd, Fe and Pb may be found in the supplementary

materials (S3–S5).

(b)(a)

Fig. 5 (a) Survival and (b) DNA damage in Corophium volutator in reference (grey bars) and contaminated sediments (black bars). Data

are mean survival (%) and DNA damage (%) ± SE (n = 6). Tukey’s post hoc comparisons contrasted survival among different levels of

pCO2 within sediment types. Bars marked with common letters are not significantly different (P > 0.05).

© 2012 Blackwell Publishing Ltd, Global Change Biology, 19, 340–351

OCEAN ACIDIFICATION AND CONTAMINATED SEDIMENT 347

sublethal effects on the test organisms. Together, these

interactions indicate a greater additive susceptibility to

both DNA damage and acute toxicity when the organ-

isms are maintained in contaminated sediments under

elevated pCO2.

Elevated pCO2 also influenced the flux of metals from

sediment bound to labile states. Ni in particular dissoci-

ated more rapidly from the test sediments at the

highest pCO2 of 1140 latm, relative to rates measured

at the background pCO2 of 390 latm (fluxes were

intermediate at 750 latm). This was observed in the

overlying OW, at the SWI and within the interstitial

PW. Surprisingly, Ni flux was not greater from the

contaminated sediment than the reference sediment,

with OA resulting in similar increased flux rates in both

sediments. A similar pattern was detected for Zn,

where fluxes of labile Zn to the PW were significantly

higher in the elevated pCO2 treatment, however, this

was only observed in the reference sediments. The fact

that fluxes were equal to, or greater, from the reference

sediment than the contaminated sediment could be

explained by differences in the sediment properties.

The contaminated sediment contained significantly

higher amounts of organic carbon and had a finer mean

particle size than the reference sediment. It is possible

that metals within the contaminated sediment were

strongly bound to fine particulates and within organic

complexes and were therefore less liable to dissociation

than metals in the sandier reference sediment.

Despite the clear effects of elevated pCO2 on the

behaviour of some metals, none of the metal fluxes

appeared to easily explain the increased toxicity at

higher pCO2. Particle-bound metals will exert toxicolog-

ical effects upon surface deposit feeding species such as

C. volutator as they ingest contaminated sediments and

may accumulate contaminants across the gut interface

(Bat & Raffaelli, 1998; Wang & Fisher, 1999). The simi-

larity in Ni flux from the reference and contaminated

sediments suggests that PW concentrations were likely

to have been the same in the two test sediments under

each level of pCO2. In combination with the fact that

metal flux for some metals was not clearly greater in

the contaminated sediments, this suggests that dietary

uptake was the predominant exposure pathway in our

experiments under all of the pCO2 treatments. DNA

damage, as assessed through the COMET assay, was

chosen as a sublethal toxicity endpoint since one of the

main mechanisms of toxicity from metals is through

oxidative damage of molecules, including DNA. The

DNA damage data suggest that the enhanced acute

mortality in the contaminated sediment at the highest

pCO2 may be driven by sublethal effects of lower pH

on C. volutator. We found significantly greater DNA

damage in the amphipods incubated in the reference

sediments at 1140 latm. The elevated pCO2, therefore,

had direct deleterious effects on the organisms. In the

contaminated sediments, the onset of DNA damage

occurred at the lower pCO2 of 750 latm. Thus, individ-

uals simultaneously exposed to contaminated sedi-

ments and OA required lower enrichment of pCO2 to

show the same effect.

A previous incubation study used passive samplers

to measure labile metal fluxes from sediments under

background and elevated pCO2 concentrations, and

found that concentrations of a wide range of labile met-

als in sediment pore water increased after exposure to

OA (Ardelan et al., 2009). These experiments were,

however, designed to examine responses to CO2 leak-

age from deep saline storage aquifers and thus tested

pCO2 concentrations much higher (up to 20 000 latm)

than those predicted to occur from OA alone. While the

levels of OA predicted to occur by the end of this

century have a strong potential to influence the specia-

tion of sediment-bound metals, not all metals can be

expected to respond similarly. Most divalent metals

form strong complexes with organic ligands and fine

particulates, and the stability of these complexes differs

between metals (Millero et al., 2009). We identified clear

effects of OA on the flux of Ni and Zn, but the majority

of metals were unaffected. Similar patterns have been

identified with water-borne metals. For example, Cu

forms strong bonds with carbonate ions in seawater,

Table 2 Nominal and calculated pCO2 and pH. Data are mean values ± SE (n = 2 samples per treatment per time)

Sediment Nominal pCO2

Calculated pCO2 pH

Day 0 Day 5 Day 9 Day 0 Day 5 Day 9

Reference 390 474.2 ± 1.7 423.6 ± 12.5 402.1 ± 34.2 8.027 ± 0.001 8.077 ± 0.008 8.084 ± 0.005

750 749.9 ± 8.2 674.9 ± 29.0 683.0 ± 18.8 7.846 ± 0.007 7.91 ± 0.017 7.885 ± 0.026

1140 1205.9 ± 4.9 1002.3 ± 43.2 965.0 ± 33.4 7.686 ± 0.013 7.748 ± 0.008 7.759 ± 0.015

Contaminated 390 464.9 ± 7.3 424.3 ± 4.5 417.5 ± 59.8 8.019 ± 0.007 8.078 ± 0.010 8.069 ± 0.016

750 749.4 ± 9.5 704.5 ± 10.0 720.5 ± 52.4 7.845 ± 0.002 7.89 ± 0.007 7.861 ± 0.001

1140 1105.3 ± 17.0 1028.3* 921.0 ± 29.5 7.647 ± 0.024 7.735 ± 0.022 7.78 ± 0.014

*Only one replicate available due to equipment failure.

© 2012 Blackwell Publishing Ltd, Global Change Biology, 19, 340–351

348 D. A. ROBERTS et al.

and under acidified conditions the free-ion Cu concen-

tration may increase as carbonate ion concentrations

decrease. In contrast, free-ion concentrations of metals

such as Cd that form pH-independent bonds may be

relatively unaffected (Pascal et al., 2010). There is a clear

need for further research examining the influence of

sediment properties such as grain size and organic

carbon content on metal fluxes under predicted future

pCO2 concentrations. This information would greatly

enhance our understanding of dissociation mecha-

nisms, as well as our ability to predict potential toxico-

logical effects of a range of sediment types. Such

studies could be conducted with artificially spiked sedi-

ments so that there are controlled metal concentrations

and parameters such as grain size can be manipulated

between experimental treatments.

Research on the effects of OA has focused on how

changes in carbonate chemistry and pH will affect rates

of calcification in marine invertebrates. OA may

also have sublethal effects on both calcifying and

noncalcifying species by affecting growth, metabolism

and reproduction (Michaelidis et al., 2005; Hauton

et al., 2009). These effects may arise as a result of dis-

ruption to acid-base regulatory mechanisms (Hauton

et al., 2009). Regulation and maintenance of stable acid-

base balance is essential to protein conformation,

enzyme function and, ultimately, metabolism. Species

that may tolerate OA will likely still divert energy away

from key biological processes, such as growth (Wood

et al., 2008; Stumpp, Wren, et al., 2011), immune func-

tion (Dupont & Thorndyke, In press) and reproduction

(Shirayama & Thornton, 2005) towards OA-compensa-

tory processes and ion homeostasis (Beniash et al.,

2010; Lannig et al., 2010; Thomsen & Melzner, 2010;

Stumpp, Dupont, et al., 2011; Stumpp,Wren, et al., 2011,

2012). According to the compensation hypothesis ani-

mals will make energetic trade-offs between different

parts of their physiological maintenance budget to meet

elevated energy demands in the face of stress. A clear

metabolic cost of resistance to metal contamination has

been demonstrated in the harbour ragworm Nereis

diversicolor living in Restronguet Creek, a highly metal

polluted estuary in Cornwall, United Kingdom (Pook

et al., 2009).

Crustaceans typically have well-developed internal

pH regulation strategies (Kroeker et al., 2010; Whiteley,

2011), with the possible exception of deep sea species

(Pane & Barry, 2007). For example, the prawns Palaemon

elegans and Palaemon serratus and crabs Necora puber

and Cancer magister may compensate for hypercapnia

(excess CO2 in blood) within days (Pane & Barry, 2007;

Spicer et al., 2007; Dissanayake et al., 2010). There are,

however, metabolic costs associated with these compen-

satory processes due to the dependence of bicarbonate

uptake from the seawater via electroneutral ion

exchange, and thus compensation may occur at the

expense of other energy-demanding processes (White-

ley, 2011). Crustaceans may also increase the transcrip-

tion of genes responsible for the production of

metabolic enzymes under OA (Hauton et al., 2009).

Cellular defence and repair processes in the face of

toxic insult, especially DNA repair systems, are also

energetically expensive processes (Deerenberg et al.,

1998). In our study, no significant induction of DNA

damage was found in the amphipods exposed to the

contaminated sediments relative to the reference sedi-

ments under background pCO2, despite the higher con-

centrations of genotoxic metals such as As, Cu and Zn.

This suggests that the suite of antioxidative defence

and DNA-repair enzymes responsible for maintaining

genetic integrity have not been overwhelmed by the

concentrations of contaminants present. However,

under a near-future pCO2 scenario of 750 latm, there is

a clear increase in DNA damage when organisms are

exposed to these contaminated sediments. Given the

absence of increases in labile metal fluxes from the con-

taminated sediments at this pCO2, it appears that the

increase in DNA damage is a result of these defence or

repair capabilities being impacted by the additional

stress of the high pCO2 conditions. In the highest pCO2

treatment of 1140 latm, animals from the ‘clean’ refer-

ence site showed significantly increased levels of DNA

damage compared with the animals exposed to the

same sediment under ambient pCO2 conditions. Very

few studies have addressed the potential for OA condi-

tions to exert oxidative stress, although this has recently

been demonstrated in the eastern oyster Crassostrea

virginica (Tomanek et al., 2011), where OA exposure

induced up-regulation of a number of antioxidant pro-

teins. Thus, the increased DNA damage observed in

C. volutator under high CO2 conditions may be due to

shifts in energy allocation between the oxidative stress

defence and repair process and the physiological costs

of pH compensation.

While we propose that the interaction between OA

and toxicity is the result of additive effects of the two

stressors (rather than synergistic effects based on effects

of reduced pH on metal speciation, for example),

further work is required to definitively elucidate the

mechanisms. Sediment properties such as carbon con-

tent and grain size are likely to influence the interaction

between sediment-bound metals and pH, and charac-

teristics of species such as their life history, behaviours

and feeding modes will influence their susceptibility to

the combined stressors. Additional experimental evi-

dence on these responses is clearly needed to broaden

our understanding of the wider implications of OA.

Regardless of the mechanisms at play, our findings do

© 2012 Blackwell Publishing Ltd, Global Change Biology, 19, 340–351

OCEAN ACIDIFICATION AND CONTAMINATED SEDIMENT 349

indicate that it may be necessary to reevaluate sediment

metals guidelines under conditions of future OA to

afford a similar level of environmental protection

to that achieved by current regulation. Our data add to

increasing evidence that environmental policy should

explicitly consider the effects of contaminants within

the context of changing background environmental

conditions which, themselves, have the potential to

exert change on benthic communities.

Acknowledgements

This work was funded by the Department for Environment,Food and Rural Affairs (Defra, contract code E5204-CHIME). C.L. is supported by a Natural Environment Research Council(NERC) UK Fellowship: NE/G014728/1. We thank StefanBolam and Alexandra Roberts for assistance in fabricatingexperimental equipment; Ruth Parker, Tom Hutchinson andSteve Widdicombe for useful discussions in the planning stage;and Robin Law, Helen Findlay, Brendan Godley and Alexan-dra Roberts for comments on earlier drafts of this manuscript.

References

Ardelan MV, Steinnes E, Lierhagen S, Linde SO (2009) Effects of experimental CO2

leakage on solubility and transport of seven trace metals in seawater and sedi-

ment. Science of the Total Environment, 407, 6255–6266.

de Backer A, van Coillie F, Montserrat F, Provoost P, van Colen C, Vincx M, Degraer

S (2011) Bioturbation effects of Corophium volutator: importance of density and

behavioural activity. Estuarine, Coastal and Shelf Science, 91, 306–313.

Bat L, Raffaelli D (1998) Sediment toxicity testing: a bioassay approach using the

amphipod Corophium volutator and the polychaete Arenicola marina. Journal of

Experimental Marine Biology and Ecology, 226, 217–239.

Beniash E, Ivanina A, Lieb NS, Kurochkin I, Sokolova IM (2010) Elevated level of car-

bon dioxide affects metabolism and shell formation in oysters Crassostrea virginica.

Marine Ecology Progress Series, 419, 95–108.

Bolam SG, Mason C, Bolam T et al. (2011) Dredged Material Disposal Monitoring Around

the Coast of England: Results of Sampling (2009). Centre for Environment, Fisheries

and Aquaculture Science, Lowestoft, UK.

Bryan GW, Langston WJ (1992) Bioavailability, accumulation and effects of heavy

metals in sediments with special reference to United Kingdom estuaries: a review.

Environmental Pollution, 76, 89–131.

Buchman M (2008) NOAA Screening Quick Reference Tables. Office of Response and

Restoration Division, National Oceanic and Atmospheric Association, Seattle,

USA.

Bulling MT, Hicks N, Murray L, Paterson DM, Raffaelli D, White PCL, Solan M

(2010) Marine biodiversity–ecosystem functions under uncertain environmental

futures. Philosophical Transactions of the Royal Society B: Biological Sciences, 365,

2107–2116.

de Deckere EMGT, van de Koppel J, Heip CHR (2000) The influence of Corophium vol-

utator abundance on resuspension. Hydrobiologia, 426, 37–42.

Deerenberg C, Overkamp GJF, Visser GH, Daan S (1998) Compensation in resting

metabolism for experimentally increased activity. Journal of Comparative Physiology

B: Biochemical, Systemic, and Environmental Physiology, 168, 507–512.

Dickson AG (1990) Thermodynamics of the dissociation of boric-acid in potassium-

chloride solutions form 273.15-K to 318.15 K. The Journal of Chemical Thermodynam-

ics, 22, 113–127.

Dickson AG, Millero FJ (1987) A comparison of the equilibrium constants for the dis-

sociation of carbonic acid in seawater media. Deep Sea Research Part A. Oceano-

graphic Research Papers, 34, 1733–1743.

Dickson AG, Sabine SL, Christian JR (2007). Guide to Best Practices for Ocean CO2 Mea-

surements. PICES Special Publication 3, IOCCP report no. 8.

Dissanayake A, Clough R, Spicer JI, Jones MB (2010) Effects of hypercapnia on acid–

base balance and osmo-/iono-regulation in prawns (Decapoda: Palaemonidae).

Aquatic Biology, 11, 27–36.

Doudican NA, Song B, Shadel GS, Doetsch PW (2005) Oxidative DNA damage causes

mitochondrial genomic instability in Saccharomyces cerevisiae. Molecular and Cellular

Biology, 25, 5196–5204.

Dupont S, Thorndyke M (In press) Relationship between CO2-driven changes in extra-

cellular acid-base balance and cellular immune response in two polar echinoderm

species. Journal of Experimental Marine Biology and Ecology, 424, 32–37.

Dupont S, Dorey N, Thorndyke M (2010) What meta-analysis can tell us about vul-

nerability of marine biodiversity to ocean acidification? Estuarine, Coastal and Shelf

Science, 89, 182–185.

Findlay HS, Wood HL, Kendall MA, Spicer JI, Twitchett RJ, Widdicombe S (2011)

Comparing the impact of high CO2 on calcium carbonate structures in different

marine organisms. Marine Biology Research, 7, 565–575.

Gattuso J-P, Hansson L (2011) Ocean Acidification. Oxford University Press, Oxford.

Hardy FG, Evans SM, Tremayne MA (1993) Long-term changes in the marine macro-

algae of three polluted estuaries in north-east England. Journal of Experimental Mar-

ine Biology and Ecology, 172, 81–92.

Hauton C, Tyrrell T, Williams J (2009) The subtle effects of sea water acidification on

the amphipod Gammarus locusta. Biogeosciences, 6, 1479–1489.

Hendriks IE, Duarte CM, Alvarez M (2010) Vulnerability of marine biodiversity to

ocean acidification: a meta-analysis. Estuarine, Coastal and Shelf Science, 86, 157–164.

Honisch B, Ridgwell A, Schmidt DN et al. (2012) The geological record of ocean acidi-

fication. Science, 335, 1058–1063.

Jha AN (2008) Ecotoxicological applications and significance of the comet assay.

Mutagenesis, 23, 207–221.

Kirkwood D (1996) Nutrients: Practical Notes on Their Determination in Seawater. Inter-

national Council for the Exploration of the Sea, Copenhagen.

Kroeker KJ, Kordas RL, Crim RN, Singh GG (2010) Meta-analysis reveals negative yet

variable effects of ocean acidification on marine organisms. Ecology Letters, 13,

1419–1434.

Lacoue-Labarthe T, Martin S, Oberhansli F, Teyssie J-L, Markich S, Ross J, Bustaman-

te P (2009) Effects of increased pCO2 and temperature on trace element (Ag, Cd

and Zn) bioaccumulation in the eggs of the common cuttlefish, Sepia officinalis. Bio-

geosciences, 6, 2561–2573.

Lacoue-Labarthe T, Reveillac E, Oberhansli F, Teyssie JL, Jeffree R, Gattuso JP (2011)

Effects of ocean acidification on trace element accumulation in the early-life stages

of squid Loligo vulgaris. Aquatic Toxicology, 105, 166–176.

Lacoue-Labarthe T, Martin S, Oberhansli F, Teyssie J-L, Jeffree R, Gattuso J-P,

Bustamante P (2012) Temperature and pCO2 effect on the bioaccumulation of

radionuclides and trace elements in the eggs of the common cuttlefish, Sepia

officinalis. Journal of Experimental Marine Biology and Ecology, 413, 45–49.

Langenbuch M, Portner HO (2002) Changes in metabolic rate and N excretion in the

marine invertebrate Sipunculus nudus under conditions of environmental hyper-

capnia identifying effective acid—base variables. Journal of Experimental Biology,

205, 1153–1160.

Langenbuch M, Portner HO (2003) Energy budget of hepatocytes from Antarctic fish

(Pachycara brachycephalum and Lepidonotothen kempi) as a function of ambient CO2:

pH-dependent limitations of cellular protein biosynthesis? Journal of Experimental

Biology, 206, 3895–3903.

Lannig G, Eilers S, Portner HO, Sokolova IM, Bock C (2010) Impact of ocean acidifi-

cation on energy metabolism of oyster, Crassostrea gigas — changes in metabolic

pathways and thermal response. Marine Drugs, 8, 2318–2339.

Lebrato M, Iglesias-Rodrıguez D, Feely RA et al. (2010) Global contribution of echino-

derms to the marine carbon cycle: CaCO3 budget and benthic compartments.

Ecological Monographs, 80, 441–467.

Lewis C, Galloway T (2008) Genotoxic damage in polychaetes: a study of species and

cell-type sensitivities. Mutation Research/Genetic Toxicology and Environmental Muta-

genesis, 654, 69–75.

Lewis E, Wallace D (1998) Program Developed for CO2 System Calculations. Carbon

Dioxide Information Analysis Center, US Department of Energy, Oak Ridge, TN.

Lewis C, Pook C, Galloway T (2008) Reproductive toxicity of the water accommo-

dated fraction (WAF) of crude oil in the polychaetes Arenicola marina (L.) and

Nereis virens (Sars). Aquatic Toxicology, 90, 73–81.

Macdonald R, Harner T, Fyfe J (2005) Recent climate change in the Arctic and its

impact on contaminant pathways and interpretation of temporal trend data. Sci-

ence of the Total Environment, 342, 5–86.

Mehrbach C, Culberson CH, Hawley JE, Pytkowicz RM (1973) Measurement of the

apparent dissociation constants of carbonic acid in seawater at atmospheric pres-

sure. Limnology and Oceanography, 18, 897–907.

Michaelidis B, Ouzounis C, Paleras A, Prtner HO (2005) Effects of long-term moder-

ate hypercapnia on acid–base balance and growth rate in marine mussels Mytilus

galloprovincialis. Marine Ecology Progress Series, 293, 109–118.

© 2012 Blackwell Publishing Ltd, Global Change Biology, 19, 340–351

350 D. A. ROBERTS et al.

Miles H, Widdicombe S, Spicer JI, Hall-Spencer J (2007) Effects of anthropogenic sea-

water acidification on acid–base balance in the sea urchin Psammechinus miliaris.

Marine Pollution Bulletin, 54, 89–96.

Millero FJ, Woosley R, Ditrolio B, Waters J (2009) Effect of ocean acidification on the

speciation of metals in seawater. Oceanography, 22, 72–85.

Morris S, Allchin CR, Zegers BN et al. (2004) Distribution and fate of HBCD and

TBBPA brominated flame retardants in North Sea estuaries and aquatic food webs.

Environmental Science and Technology, 38, 5497–5504.

Neuparth T, Correia AD, Costa FO, Lima G, Costa MH (2005) Multi-level assessment

of chronic toxicity of estuarine sediments with the amphipod Gammarus locusta:

I. Biochemical endpoints.Marine Environmental Research, 60, 69–91.

Noyes PD, McElwee MK, Miller HD et al. (2009) The toxicology of climate change:

environmental contaminants in a warming world. Environment International, 35,

971–986.

Pane EF, Barry JP (2007) Extracellular acid–base regulation during short-term hyper-

capnia is effective in a shallow-water crab, but ineffective in a deep-sea crab.

Marine Ecology Progress Series, 334, 1–9.

Pascal P-Y, Fleeger JW, Galvez F, Carman KR (2010) The toxicological interaction

between ocean acidity and metals in coastal meiobenthic copepods. Marine Pollu-

tion Bulletin, 60, 2201–2208.

Plater A, Ridgway J, Appleby P, Berry A, Wright M (1999) Historical contaminant

fluxes in the Tees Estuary, UK: geochemical, magnetic and radionuclide evidence.

Marine Pollution Bulletin, 37, 343–360.

Pook C, Lewis C, Galloway T (2009) The metabolic and fitness costs associated with

metal resistance in Nereis diversicolor. Marine Pollution Bulletin, 58, 1063–1071.

Quinn GP, Keough MJ (2002) Experimental Design and Data Analysis for Biologists.

Cambridge University Press, Cambridge.

Richards R, Chaloupka M, Sano M, Tomlinson R (2011) Modelling the effects of

“coastal” acidification on copper speciation. Ecological Modelling, 222, 3559–3567.

Riebesell U, Fabry VJ, Hansson L, Gattuso J-P (2010) Guide to Best Practices for Ocean

Acidification Research and Data Reporting. Publications Office of the European

Union, Luxembourg.

Roddie B, Thain J (2001) Biological Effects of Contaminants: Corophium sp. Sediment Bio-

assay and Toxicity Test. ICES, Copenhagen.

Rogers AD, Laffoley D (2011) International Earth System Expert Workshop on Ocean

Stresses and Impacts. Summary Report. International Program on the State of the

Ocean, Oxford.

Sabine CL, Freely RA, Gruber N et al. (2004) The oceanic sink for anthropogenic CO2.

Science, 305, 367–371.

Shirayama Y, Thornton H (2005) Effect of increased atmospheric CO2 on shallow

water marine benthos. Journal of Geophysical Research, 110, C09S08.

Royal Society R (2005) Ocean Acidification Due to Increasing Atmospheric Carbon Dioxide.

The Royal Society, London.

Spicer J, Raffo A, Widdicombe S (2007) Influence of CO2-related seawater acidificat-

ion on extracellular acid–base balance in the velvet swimming crab Necora puber.

Marine Biology, 151, 1117–1125.

Spicer J, Widdicombe S, NeedhamHR, Berge JA (2011) Impact of CO2-acidified seawa-

ter on the extracellular acid-base balance on the northern sea urchin Strongylocentro-

tus droebachiensis. Journal of Experimental Marine Biology and Ecology, 407, 19–25.

Stohs SJ, Bagchi D (1995) Oxidative mechanisms in the toxicity of metal ions. Free

Radical Biology and Medicine, 18, 321–336.

Stumpp M, Dupont S, Thorndyke MC, Melzner F (2011) CO2 induced seawater

acidification impacts sea urchin larval development II: gene expression patterns in

pluteus larvae. Comparative Biochemistry and Physiology – Part A: Molecular & Inte-

grative Physiology, 160, 320–330.

Stumpp M, Wren J, Melzner F, Thorndyke MC, Dupont ST (2011) CO2 induced

seawater acidification impacts sea urchin larval development I: elevated meta-

bolic rates decrease scope for growth and induce developmental delay. Com-

parative Biochemistry and Physiology – Part A: Molecular & Integrative Physiology,

160, 331–340.

Stumpp M, Trubenbach K, Brennecke D, Hu MY, Melzner F (2012) Resource

allocation and extracellular acid–base status in the sea urchin Strongylocentrotus

droebachiensis in response to CO2 induced seawater acidification. Aquatic Toxicol-

ogy, 110–111, 194–207.

Thomsen J, Melzner F (2010) Moderate seawater acidification does not elicit long-

term metabolic depression in the blue mussel Mytilus edulis. Marine Biology, 157,

2667–2676.

Tomanek L, Zuzow MJ, Ivanina AV, Beniash E, Sokolova IM (2011) Proteomic

response to elevated pCO2 level in eastern oysters, Crassostrea virginica: evidence

for oxidative stress. The Journal of Experimental Biology, 214, 1836–1844.

Turley C, Eby M, Ridgwell AJ et al. (2010) The societal challenge of ocean acidificat-

ion. Marine Pollution Bulletin, 60, 787–792.

Underwood AJ (1996) Experiments in Ecology: Their Logical Design and Interpretation

Using Analysis of Variance. Cambridge University Press, Cambridge.

Valko M, Morris H, Cronin MTD (2005) Metals, toxicity and oxidative stress. Current

Medicinal Chemistry, 12, 1161–1208.

Wang W-X, Fisher NS (1999) Delineating metal accumulation pathways for marine

invertebrates. Science of the Total Environment, 237–238, 459–472.

Whiteley NM (2011) Physiological and ecological responses of crustaceans to ocean

acidification. Marine Ecology Progress Series, 430, 257–271.

Wood HL, Spicer JI, Widdicombe S (2008) Ocean acidification may increase calcifica-

tion rates, but at a cost. Proceedings of the Royal Society B: Biological Sciences, 275,

1767–1773.

Supporting Information

Additional Supporting Information may be found in theonline version of this article:

Table S1. Analysis of variance (ANOVA) for metal concentra-tions in overlying water (OW) and sediment–water interface(SWI) DGTs.Table S2. Analysis of variance (ANOVA) for metal concentra-tions in the sediment DGTs. Bold values are statistically sig-nificant (P < 0.05).Table S3. Post hoc comparisons for significant ‘sedi-ment 9 pCO2’ interactions for metal concentrations in thesediment DGTs.Table S4. Post hoc comparisons for significant ‘sedi-ment 9 depth’ interactions for metal concentrations in thesediment DGTs.Figure S5. Depth profiles of labile Cd, Fe and Pb flux fromreference and contaminated sediments to the pore water.Data are mean metal fluxes (nmol cm�1 sec�1) ± SE (n = 2).Mean metal concentrations in the DGT blanks are depictedas a cross on the x-axis for comparison.

© 2012 Blackwell Publishing Ltd, Global Change Biology, 19, 340–351

OCEAN ACIDIFICATION AND CONTAMINATED SEDIMENT 351