MOF, Alm.del - 2019-20 - Endeligt svar på spørgsmål 781: Spm. om, hvilke undersøgelser af sediment og biota der er foretaget i det område, som påvirkes af forureningen af farlige stoffer fra Rønland (Cheminova-jordforurening) med henblik på at konstatere forureningens udbredelse i Limfjorden, til miljøministeren

Miljø- og Fødevareudvalget 2019-20
MOF Alm.del
Offentligt

Environmental Pollution xxx (xxxx) 113985

Contents lists available at ScienceDirect

Environmental Pollution

Elevated mercury concentrations in biota despite reduced sediment concentrations in a

contaminated coastal area, Harboøre Tange, Denmark

☆

Poul Bjerregaard

∗

, Torben Grau Schmidt, Maria Pedersen Mose

Department of Biology, University of Southern Denmark, Campusvej 55, DK-5230 Odense, Denmark

ARTICLE INFO

Article history:

Received 7 October 2019

Received in revised form 8 January 2020

Accepted 13 January 2020

Available online xxx

ABSTRACT

Keywords

Mercury

Sediment

Biota

Benthic

Metals sequestered in coastal sediments are normally considered to be stable, but this investigation shows

–

somewhat surprisingly

–

that mercury concentrations in a previously contaminated area, Harboøre Tange, Den-

mark, have decreased since the 1980s. Mercury concentrations were determined in sediment and benthic biota

and present values were compared to values in the 1980s and values from areas without known; history of mer-

cury contamination. Concentrations in both the upper 20 cm of the sediments and; biota are considerably lower

now compared to latest monitoring (1980s). Sediment.concentrations at most locations have decreased from the

100–300 ng Hg g

−1

dry weight (dw) level to levels below the Background Concentration (BC) of 50 ng Hg g

−1

dw defined by Oslo-Paris Convention for the Protection of the Marine Environment of the North-East Atlantic;

some stations are at the 2–10 ng Hg g

−1

dw level characteristic of Danish coastal sediments with no known

history of mercury contamination. Concentrations of mercury in the benthic biota along Harboøre Tange have

also decreased since the 1980s but despite the lowered mercury concentrations in the sediments, concentrations

in most samples of benthic invertebrate fauna still exceed those in uncontaminated coastal areas and also the

Environmental Quality Standard (EQS) of 20 ng Hg g

−1

wet weight (≈100 ng Hg g

−1

dry weight) defined by

the European Union's Water Framework Directive. Concentration ranges in selected organisms are: (Harboøre

Tange l980s/Harboøre Tange now/uncontaminated areas - given in ng Hg g

−1

dw): Periwinkles

Littorina littorea

9000/150–450/55-77, blue mussels

Mytilus edulis

up to 9000/300–500/40–170, cockles

Cerastoderma edule

to 8000/400–1200/200, brown shrimp

Crangon crangon

700–2200/150-450/47, eelgrass

Zostera marina

up to

330/25–70/12. The present results - together with a literature review - show that a simple and straight forward

relationship between the concentrations of mercury in sediment and benthic organisms does not necessarily exist.

1. Introduction

Coastal and estuarine sediments have long been recognized as sinks

for pollutants, and once buried in the sediments, metals and non-degrad-

able, organic contaminants may persist for years or decades (e.g.

Baeyens et al., 2005; Ridgway and Shimmield, 2002). Due to its ca-

pacity to be methylated and accumulate to harmful concentrations along

aquatic food chains (Grandjean et al., 1997; Scheuhammer et al.,

2008), releases of mercury to the environment from point sources and

on a more global scale have been a major concern for decades (UNEP,

2013).

The western part of Limfjorden, Denmark, Nissum Broad along

Harboøre Tange (Fig. 1) was heavily contaminated with mercury dur-

ing the 1950s and 1960s - mainly due to discharges from the pesti-

cide-producing factory, Cheminova (Lyngby and Brix, 1987). It is esti

∗

☆

This paper has been recommended for acceptance by Prof. Wen-Xiong Wang.

Corresponding author.

E-mail address:

[email protected] (P. Bjerregaard)

mated that approximately 30 tonnes of mercury were discharged into

the Nissum Broad with process waste water, lost by production mistakes

and deposited in the vicinity of the factory (Kiorboe et al., 1983).

Thresholds and quality criteria for mercury in sediment and biota

have been defined by various organizations. The Background Concen-

tration (BC) and Background Assessment Concentration (BAC) for mer-

cury in sediment are 50 and 70 ng Hg g

−1

dw, respectively, defined

by the Oslo-Paris Convention for the Protection of the Marine Envi-

ronment of the North-East Atlantic (OSPAR, 2009). The threshold be-

low which there is a low risk of adverse effects towards the biota (Ef-

fects Range Low

–

ERL) is 150 ng Hg g

−1

dw sediment (Long et al.,

1998; Long et al., 1995; OSPAR, 2009). The Environmental Qual-

ity Standard (EQS) of 20 ng Hg g

−1

wet weight (corresponding to ap-

proximately 100 ng Hg g

−1

dw) in invertebrates and fish defined in the

European Union Water Framework Directive (EC, 2008; EU, 2008) is

supposed to protect the marine food chains from biomagnifying mer-

cury to harmful concentrations in top predators; the potential toxic-

ity associated with elevated levels of mercury in aquatic food chains

is well documented (reviewed by Driscoll et al., 2013). Until 2013

https://doi.org/10.1016/j.envpol.2020.113985

journal homepage: http://ees.elsevier.com

MOF, Alm.del - 2019-20 - Endeligt svar på spørgsmål 781: Spm. om, hvilke undersøgelser af sediment og biota der er foretaget i det område, som påvirkes af forureningen af farlige stoffer fra Rønland (Cheminova-jordforurening) med henblik på at konstatere forureningens udbredelse i Limfjorden, til miljøministeren

P. Bjerregaard et al. / Environmental Pollution xxx (xxxx) 113985

Fig. 1.

Maps showing the sampling sites. The left insert shows Harboøre Tange with the sampling stations 1–10 and N and S. The positions of the old, demolished and the present chemical

factories are indicated with a grey dot and circle, respectively. The right insert shows the island of Funen with the reference stations.

2. Materials and methods

2.1. Field sampling

Sediment samples were collected at the 10 stations along Harboøre

Tange indicated in Fig. 1 on October 30, 2014. Stations 1 to 9 are sit-

uated at open, coastal locations whereas station 10 is located in a se-

cluded cove with less water exchange and wave action than at the other

stations. Triplicate cores were taken at each station by means of Kajak

samplers (50 cm polystyrene tube with a diameter of 5.2 cm) except at

station 10 where 6 cores were taken because of the expected higher di-

versity of the sediment in the secluded cove; depth of the individual

cores varied between 10 and 24 cm. At station 4, the stony character

of the sediment prevented sampling of cores. Further details on the sed-

iment sampling are given in Supplementary Material, Table S1. The

method of sediment sampling was the same as used by Lyngby and

Brix (1987) with which comparisons are made.

Specimens of the biota in an area up to 15 m from the site of the sed-

iment sample were collected by hand and net; a radius up to 15

the Danish environmental authorities used a Norwegian Environmental

Classification System (NECS) to assess conditions regarding mercury in

the coastal areas (Bakke et al., 2007); NECS focusses on data from sed-

iment, blue mussels

Mytilus edulis,

periwinkles

Littorina littorea

and rock

weed

Fucus vesiculosus.

Investigations during the 1980s documented highly elevated concen-

trations of Hg in sediment and biota along Harboøre Tange (Andersen,

1992; Brix and Lyngby, 1984; Kiorboe et al., 1983; Lyngby and

Brix, 1987; Riisgard, 1984); concentrations in sediment with up to

μg

Hg g

−1

dw exceeded the ERL and natural background concentra-

tions by more than 3 orders of magnitude at the most contaminated site;

eelgrass (up to 1130 ng Hg g

−1

dw) and blue mussels (up to 15

μg

−1

dw) exceeded natural background concentrations by up to two or-

ders of magnitude (Lyngby and Brix, 1987). Since then, mercury con-

centrations in sediments and biota in the area have not been monitored.

With the hypothesis that mercury concentrations had remained con-

stant since the 1980s, the main purpose of this investigation was to as-

sess mercury concentrations in sediment and benthic biota in the pre-

viously contaminated area and compare these values to previous values

from the area and to determine mercury concentrations in areas without

known history of mercury contamination. Both sets of data are evalu-

ated against the various thresholds defined in legislations and conven-

tions.

m was used to secure enough organisms. If abundant, 5 specimens of

each species were collected

–

otherwise 1 to 4 specimens were collected.

Leaves were taken from eelgrass,

Zostera marina,

and top shoots were

taken from the macroalgae; the benthic fauna was collected as intact or-

ganisms.

On February 4, 2015 similar sampling was carried out at the two ref-

erence sites Enebærodde, Funen, Denmark (indicated in Fig. 1, R1 &

R2)

–

without any known history of mercury contamination.

Because the NECS focus species (blue mussels, periwinkles and rock

weed) were sparsely abundant at the two reference sites (R1 and R2) it

was decided to further characterise background concentrations of mer-

cury in the in areas with no known history of mercury contamination.

Benthic biota was sampled twice at 4 reference sites (B: Ballen, F: Fald-

sled K: Kerteminde and N: Nyborg) around the island of Funen (Fig. 1)

during May (B1, F1, K1, N1) and September–October (B2, F2, K2, N2)

2016. Five and approximately 20 specimens of each species were col-

lected from each site in May and September–October, respectively. One

further reference site (Bo: Bogense, Fig. 1) was sampled June 3, 2016.

Biota was further collected at two sites along Harboøre Tange (S:

Harboøre South and N: Harboøre North, Fig. 1) on June 11, 2016. Ref-

erence shrimps were collected at Kerteminde, July 17, 2017 (K3).

Further details on sampling dates, numbers and sizes of specimens

collected at the various sampling sites are given in Supplementary Ma-

terial, Tables S2–S5.

All samples

–

biota and sediment cores

–

were collected at walking

depths close to the shore at water depths of 20–100 cm. Coordinates for

the sampling locations are given in Table S6.

2.2. Treatment of samples

2.2.1. Sediment

The sediment cores were split into 2 cm fractions that were weighed,

frozen, freeze dried and weighed again to determine the water content.

A subsample of the freeze-dried sediment was heated to 500 °C for 5 h

to determine the organic content (Loss on Ignition

–LOI).

Another sub-

sample of the freeze-dried sediment was used for mercury analysis.

2.2.2. Biota

The biota was transported the 250 km to the laboratory in buck-

ets with aerated water and processed the day after the sampling. The

size of the mollusks and shore crabs was determined by measuring

the height, width or length of their shell or carapace. The other ani

P. Bjerregaard et al. / Environmental Pollution xxx (xxxx) 113985

mals were weighed. No size estimate was made for plants and macro

algae since only the top shoots were collected. The soft parts were dis-

sected from the mollusks and the shore crabs; shrimps and other animals

were treated as whole organisms. The samples were frozen and subse-

quently freeze dried.

2.3. Mercury analysis

Total mercury was determined by means of a Milestone DMA-80 Di-

rect Mercury Analyser. The quality of the determinations was validated

by the use of PACS-1 and PACS-3 sediment standards and TORT-stan-

dards (lobster hepatopancreas) with certified mercury contents; blanks

were included. Values were within ±10% of the certified values. Up to

50 mg dry weight sample was used in the analysis. The mercury concen-

tration was determined in each individual specimen. All concentrations

are given on dry weight (dw) basis; sediment concentrations are not nor-

malised with regard to organic content (but data on this are available

in Supplementary material, Table S1). This method detects HgS in the

anaerobic compartment of the sediment; HgS decomposes at tempera-

tures of 3–400 °C (Azzaria and Aftabi, 1991) - well below the 650 °C

the samples are heated to during the analysis.

2.4. Data handling and statistics

Analyses of variance were used to identify differences between sam-

pling sites and times; a few data sets had to be log

-transformed to ob-

tain normal distributions. Tukey's posttest was used for pairwise com-

parisons. The results of the statistical analyses are indicated in the fig-

ures; data points with no common lower-case letter are significantly dif-

ferent. Pearson correlation analyses were used to identify relations be-

tween parameters. 0.05 was used as significance level. SYSTAT© ver. 13

was used in the statistical analyses. Because raw data from the 1980s′

investigations were not available it was not possible to perform a formal

statistical comparison.

3. Results

3.1. Sediments

Mercury concentrations in the sediment at each station are reported

to the depth of the shortest of the individual cores in Fig. 2; the content

of water and organic material (LOI) is given in Supplementary Mater-

ial Table S1 together with the mercury concentration based on the or-

ganic content.

Most of mercury concentrations in the sediment at the reference sites

are in the range 2–5 ng Hg g

−1

dw with no apparent trends in the depth

profile except for an unusually high value at R2 in 4 cm's depth (Fig.

2). At Harboøre Tange stations 1, 6, 7, and 8, Hg concentrations in the

sediment were close to background levels and generally below 10 ng Hg

−1

dw. Stations 2, 3, 5, and 9 had sediment concentrations between 20

and 40 ng Hg g

−1

dw while sediment concentrations at station 10 were

highly variable with values in individual core sections ranging between

5 and 10,800 ng Hg g

−1

dw and an average of 592 ng Hg g

−1

in all of the

core sections. For all of the stations with elevated mercury concentra-

tions, mercury appeared to be almost evenly distributed along the depth

profiles (Fig. 2).

At most of the stations where comparisons with earlier reported val-

ues were possible, mercury concentrations in the upper 5–6 cm of the

sediment had decreased considerably since the 1980s (Fig. 3).

3.2. Biota

Besides the 3 NECS focus species (blue mussels, periwinkles and

rock weed), 6 additional species were present at a sufficiently high

number of sampling stations to allow meaningful comparisons.

Fig. 2.

Depth profiles for mercury from the 2014 sampling. Sampling locations 1–3, 5–10,

R1 and R2 are shown in Fig. 1.

These were the cockle

Cerastoderma edule,

three decapod crustaceans

(the shore crab

Carcinus maenas,

the brown shrimp

Crangon crangon

and

the grass prawn

Palaemon elegans),

eelgrass

Zostera marina

and the slimy

whip weed

Chordaria flagelliformis.

These 9 species are termed the main

species of the present investigation.

3.2.1. Mollusks

At the 2016 Funen reference sites, mercury concentrations in blue

mussels and periwinkles did not differ significantly between the May

and September–October samplings and the combined results are shown

in Fig. 4AB.

Mercury concentrations in the mussels from the Funen reference

sites correlated positively with the size of the animal (length: r = 0.53,

p < 0.001). Average mercury concentrations in blue mussels at B, Bo,

K and N were in the range 39–65 ng Hg g

−1

dw whereas average con-

centrations at F and R1 were in the range 122–170 ng Hg g

−1

dw and

significantly higher (Fig. 4A); mussels collected at F were larger than at

the other sites, explaining the statistically significant difference in mer-

cury concentrations; however, this did not explain the higher values at

R1. At the Harboøre Tange sites average mercury concentrations in the

range 315–499 ng Hg g

−1

dw were all significantly higher than at the

Funen reference sites (Fig. 4A).

Average mercury concentrations in the periwinkles at the reference

sites ranged between 55 and 77 ng Hg g

−1

dw with no significant dif-

ference between the stations (Fig. 4B). There was no statistically sig-

nificant correlation between the size of the periwinkles and their mer-

cury concentration. Concentrations along Harboøre Tange were signifi

P. Bjerregaard et al. / Environmental Pollution xxx (xxxx) 113985

Fig. 3.

Mercury concentrations in the upper 5–6 cm of the sediments from Harbøre Tange

( ) and reference ( ) stations from the 2014 sampling; mean ± SEM, n = 3. Stations

with no common lower-case letter show statistically significant (p < 0.05) differences. For

comparison, mercury concentrations at the same locations in 1983 (Brix and Lyngby,

1984) are shown ( ). Sampling locations are shown in Fig. 1. (For interpretation of the

references to color in this figure legend, the reader is referred to the Web version of this

article.)

cantly higher at all sites than at the reference sites with average values

ranging between 155 and 443 ng Hg g

−1

dw.

Cockles at the reference site generally showed higher mercury con-

centrations (201 ng Hg g

−1

dw) than the other two mollusc species and

the concentrations at stations

–

especially along the southern part of

- Harboøre Tange were generally higher than at the reference station

(Fig. 4C).

No statistically significant correlations were found between the mer-

cury concentrations in the surface (upper 6 cm) sediments at the indi-

vidual sampling stations along Harboøre Tange and the concentrations

in the mollusks; this was true both when the sediment mercury concen-

trations were expressed on basis of dry weight and organic content.

3.2.2. Decapod crustaceans

C. maenas

at the reference site had mercury concentrations of

48 ± 3 ng Hg g

−1

dw in their soft parts (Fig. 5A) and average val-

ues for

C. crangon

and

P. elegans

at reference site K3 were 47 ± 3 and

66 ± 23 ng Hg g

−1

dw, respectively (Fig. 5BC) with

P. elegans

showing

a high variability compared to

C. crangon.

Average concentrations at the

sites along Harboøre Tange were up to an order of magnitude higher

than at the reference stations but for the shrimps, the difference was not

statistically significant at all of the stations (Fig. 5BC)

–

probably due to

the low number of individuals.

No statistically significant correlations were found between the mer-

cury concentrations in the surface (upper 6 cm) sediments at the indi-

vidual sampling stations along Harboøre Tange and the concentrations

in the crustaceans.

3.2.3. Plants

Eelgrass was not found at any of the Funen reference stations. Mer-

cury concentrations in the eelgrass leaves at the Harboøre Tange sites

appeared higher (Fig. 6A) than the background values of 12 ng Hg g

−1

dw for the Limfjord generally (Lyngby and Brix, 1987).

4. Discussion

Fig. 4.

Mercury concentrations (mean ± SEM; n shown in Tables S2 and S3) in three

molluscs along Harboøre Tange ( : 2014; crosshatched: 2016) and at the Funen reference

sites (

). Stations with no common lower-case letter show statistically significant

(p < 0.05) differences. Literature values from the 1980s are indicated on the Y axis to the

right. Blue mussels: (Brix and Lyngby, 1984) ( ) and (Riisgard, 1984) ( ); periwin-

kles: (Kiorboe et al., 1983) ( ); Cockles (Kiorboe et al., 1983) ( ) and (Mohlenberg

and Riisgard, 1988) ( ). The Environmental Quality Standard is indicated by the dot-

ted line. (For interpretation of the references to color in this figure legend, the reader is

referred to the Web version of this article.)

Mercury concentrations in the eelgrass leaves at Harboøre Tange

showed a positive correlation (p = 0.005) with sediment mercury con-

centrations; the positive correlation was mainly driven by the high con-

centrations at station 10.

Average background concentrations of mercury in the seaweeds were

approximately 10 ng Hg g

−1

dw with little variability (Fig. 6BC). The

concentrations at the Harboøre Tange sites were generally two-to

five-fold higher (Fig. 6BC).

3.2.4. Less abundant organisms

Concentrations of mercury in less abundant species are presented in

Tables S4 & S5 and in Figs. S1–3. For these organisms, the limited

overlap between presence at reference and Harboøre Tange sites does

not allow firm conclusions to be drawn; however, judged from back-

ground concentrations extracted from the scientific literature (Table

S7), concentrations in some of these organisms along Harboøre Tange

do appear to be elevated.

The overall conclusion from this investigation is that mercury con-

centrations in the sediment along Harboøre Tange have decreased con-

siderably since the 1980s, but also that the mercury concentrations

in the benthic invertebrates are higher than background levels and

P. Bjerregaard et al. / Environmental Pollution xxx (xxxx) 113985

Fig. 5.

Mercury concentrations in three species of decapod crustaceans along Harboøre

Tange ( : 2014; crosshatched: 2016) and at a Funen reference site ( ). Literature val-

ues from the 1980s are indicated on the Y axis to the right for

C. crangon

(Riisgard and

Famme, 1986) (■) and (Kiorboe et al., 1983) ( ). Symbols as in Fig. 4. (For inter-

pretation of the references to color in this figure legend, the reader is referred to the Web

version of this article.)

still exceed the EQS, despite sediment concentrations generally being be-

low the BAC and BC values defined by OSPAR.

4.1. Mercury in sediment

Sediment mercury concentrations determined at the reference sta-

tions (generally < 10 ng g

−1

) agreed well with literature values

(6–13 ng g

−1

) obtained in the Baltic area (e.g. Beldowski et al., 2014)

–

as such generally much lower than OSPAR's BC/BAC values of 50/

70 ng g

−1

(OSPAR, 2009).

At Harboøre Tange, mercury concentrations in the sediment only ex-

ceed the ERL value in the secluded cove at station 10, indicating that at

the other stations along the coast there is little risk that the remaining

mercury might lead to adverse effects in the biota.

Several investigations have shown that mercury deposited in coastal

sediments due to discharges from point sources has the potential to

spread.

After dredging of the most polluted sediment (>25

μg

Hg g

−1

) in

the Minamata Bay in 1990, the concentrations of total mercury in the

surface sediments of the highly polluted bay were reported by the Ku-

mamoto Prefecture (2010) (cited by Matsuyama et al., 2016) to

average 4.1

μg

Hg g

−1

. An investigation in 2012 showed that the con-

centrations changed somewhat during the preceding 25 years where

the average concentration of total mercury was determined to 3.0

μg

Hg g

−1

in the total sediment and 2.3

μg

Hg g

−1

in the surface of

Fig. 6.

Mercury concentrations in eelgrass and two seaweeds along Harboøre Tange ( :

2014; crosshatched: 2016) and at Funen reference sites ( & ). Literature values from

the 1980s are indicated on the Y axis to the right for eelgrass ( ) (Brix and Lyngby,

1984). BG is background values from Brix and Lyngby (1984). Symbols as in Fig. 4.

(For interpretation of the references to color in this figure legend, the reader is referred to

the Web version of this article.)

the sediment (Akito et al., 2014). Tomiyasu et al. (2014) found no

significant variation in sediment samples from 2002, 2006, 2008 and

2010 with values in the range of 2.47 and 3.34

μg

Hg g

−1

each of the

years. Although a relatively small change of the sediment concentrations

have been seen after the dredging, transport of mercury from the Mina-

mata Bay to the Yatsushiro Sea has been documented (Balogh et al.,

2015). The transport was suggested to be caused by a slow (~110 m/y)

migration of mercury-bearing sediment particles (Kudo et al., 2000).

The annual particulate transport of total mercury from the Minamata

Bay to Yatsushiro Sea has been estimated to 6 kg (Yano, 2013) and

modelled to150 kg (Rajar et al., 2004) hereof 16 kg under normal

and 132 kg under storm conditions. For comparison the total amount

of total mercury mobilized from the sediment in the Minamata Bay to

the water column was estimated to 0.7 kg (Akito et al., 2014). This

agrees with results on mercury transport in the contaminated Gulf of

Trieste where model considerations and actual measurements showed

that most mercury is transported adsorbed on suspended particles and

that transport of dissolved total mercury is almost negligible (Rajar et

al., 2004). Similarly, mercury discharged to the Bay of Kuwait during

1963–1985 appears to be spreading towards the northern part of the

Bay (Al-Zamel et al., 2010). Mercury concentrations in the top 2 cm

of the sediments outside a former chlor-alkali plant in Chaleur Bay, New

Brunswick, Canada have decreased considerably after the plant closed

in 2008 (Walker, 2016) and the author ascribes this decrease to burial

by recent sediment deposition.

P. Bjerregaard et al. / Environmental Pollution xxx (xxxx) 113985

Mercury concentrations in the sediment along Harboøre Tange ap-

pear to have decreased at a faster rate and to lower concentrations than

seen in the examples mentioned above. It is not known with certainty

which process underlies the decrease in mercury concentrations and

where the mercury has gone but 3 explanations have been considered:

Firstly, the mercury might now be buried deeper in the sediment

than our Kajak samplers reached. Mercury in the Minamata Bay does not

appear to move downwards in the sediments (Tomiyasu et al., 2006)

and there is nothing in the depth profiles we have seen at Harboøre

Tange suggesting that higher mercury concentrations should be found

below our greatest depths for sampling, but the possibility cannot be

completely rejected.

Secondly, there might be a possibility that the divalent mercury has

been reduced to elemental mercury and evaporated as such. It is known

from studies on River Elbe's mercury-contaminated floodplains that di-

valent mercury is reduced to elemental mercury that may evaporate to

the atmosphere (During et al., 2009; Rinklebe et al., 2010; Rin-

klebe et al., 2009; Rinklebe et al., 2013). Reduction of Hg

to Hg°

- mediated by microorganisms - takes place in the free water phase in

the marine environment (Monperrus et al., 2007) and this process has

also been demonstrated to take place in the sediment in shallow, coastal

areas. The values found for the release of elemental mercury from sed-

iment water to the atmosphere range from 0.04 pg Hg m

−2

hour

−1

the St. Lawrence River, Canada (Poissant et al., 2007), to over 80 ng

Hg m

−2

hour

−1

in a mercury contaminated lagoon in Italy (Covelli et

al., 2008; Emili et al., 2012). In the tidal environment in Arcachon

Bay, France, Bouchet et al. (2011) reported sediment-water fluxes up

to 5.1 ng Hg m

−2

hour

−1

, sediment-atmosphere fluxes up to 40 ng Hg

−2

hour

−1

and water-atmosphere fluxes up to 14.5 ng Hg m

−2

hour

−1

Sharif et al. (2013) and Conaway et al. (2003) reported fluxes up to

15 ng Hg m

−2

hour

−1

in the Gironde Estuary, France, and up to 45 ng Hg

−2

hour

−1

in the San Francisco Bay, respectively. It has also been found

that e.g. eelgrass can absorb mercury from the sediment and release it to

the atmosphere as elemental mercury; rates of 10–40 ng Hg m

−2

hour

−1

have been reported (Lindberg et al., 2005; Lindberg et al., 2002)

and Bouchet et al. (2011) observed that the presence of eelgrass in

the Arcachon Bay increased the mercury fluxes under light conditions.

For reduction and evaporation of divalent mercury to explain the de-

crease in sediment concentrations along Harboøre Tange, calculations

show that a flux rate of at least 200 ng Hg m

−2

hour

−1

over the 32 years

between 1982 (Lyngby and Brix, 1987) and 2014 would have been re-

quired. Preliminary experiments with sediment cores from station 10 of

the present experiment showed that the flux (with 2–4 cm water phase)

to the atmosphere was around 3.4 ng Hg m

−2

hour

−1

(Schultz, 2017)

–

or far less than needed to explain the reduction in the sediment mercury

concentrations over the 3 decades.

The third possibility is that mercury has been removed with trans-

port of sediment and water. The water currents in Nissum Broad are

mainly driven by the 30–40 cm tidal changes in the North Sea at this

location and there is a net flow in the Limfjord from west to east of

300–400 m

−1

but the actual inflow and outflow from and to the

North Sea are approximately 10 times higher (DTU-Aqua, 2017). All

of the sampling stations of the present investigation were situated in

areas with shallow water affected by wave action. It is possible that

wave action may bring especially the minor organic particulates with

the highest mercury content (Kudo et al., 2000; Rajar et al., 2004;

Yano, 2013) into suspension whereby the water currents may have

transported the mercury both to the North Sea and eastward in the

Limfjord. This is probably the most plausible explanation for the de-

crease in the sediment concentrations along Harboøre Tange; this hy-

pothesis also appears to be corroborated by the fact that sediment mer-

cury concentrations have remained high at station 10, situated in a se-

cluded cove with less wave action than at the other stations. Confirma

tion of this hypothesis would, however, require an intensive monitoring

programme.

4.2. Mercury in biota

At most of the reference stations, the mercury concentrations in the

soft parts of the blue mussels are in the range of

‘low

concentrations’

(50 ng Hg g

−1

dw) defined by OSPAR (2009) and they rarely exceed

OSPAR's (2009) BAC (90 ng Hg g

−1

dw). Judged from the average Hg

concentrations in soft parts of blue mussels, the NECS classification sys-

tem (Bakke et al., 2007) would characterise the area along Harboøre

Tange as

‘moderately

contaminated’ (values between 200 and 500 ng

Hg g

−1

dw), whereas the values for periwinkles and rock weed would

result in a characterisation as

‘uncontaminated’

(below 500 and 50 ng

Hg g

−1

dw for the periwinkles and rock weed, respectively). It is obvi-

ous that the NECS thresholds for the classification

‘uncontaminated’

are

well above the background levels at the Danish reference sites for all of

these 3 focus species. It is also apparent that all of the benthic inverte-

brates collected along Harboøre Tange have mercury concentrations ex-

ceeding the EQS of approximately 100 ng Hg g

−1

–

and thereby also

exceeding the blue mussels BC and BAC values of 50 and 90 ng Hg g

−1

dw. Although mercury concentrations in the present-day benthic fauna

along Harboøre Tange still exceed the EQS, BC and BAC values it is ap-

parent that a major decrease has taken place since the 1980s.

4.3. Relationship between concentrations of mercury in sediment and biota

Although mercury concentrations in the surface sediments at 9 out

of the 10 stations along Harboøre Tange can be characterized as back-

ground according to both OSPAR and NECS, the benthic fauna in the

area show elevated mercury levels.

Most data on the relationship between concentrations of mercury in

sediment and benthic organisms exists for various species of blue mus-

sels

Mytilus

sp. Both inorganic and organic mercury in the sediments

have been shown in laboratory experiments to be available for uptake in

blue mussels (Gagnon and Fisher, 1997) but field investigations show

variable results.

Clear correlations between sediment mercury concentrations in the

Limfjord area and mercury concentrations in blue mussels were identi-

fied by Lyngby and Brix (1987) and statistically significant correla-

tions were also found in investigations in Galicia, Spain (Beiras et al.,

2003; Beiras et al., 2002) and Narragansett Bay, New England, USA

for bivalves (including

M. edulis)

(Taylor et al., 2012) whereas Spada

et al. (2012) only saw a trend (p = 0.067

–

calculated from data in

their Table 2) in the Taranto Bay, Italy. Correlation was not found for

blue mussels in the present investigation

–

probably caused by the low

number of sites where blue mussels were found; failure to demonstrate

statistically significant correlations in areas in which such is actually

present may be caused by too limited sample size or too low variability

in the mercury levels.

Also for the other benthic animals no such correlations were found

in the present investigation and judged from the information present in

the scientific literature presented below, the relationship between sedi-

ment and animal concentrations does not appear to be straight forward.

Periwinkles were obtained from 9 sampling stations in the present

investigation with surface sediment concentrations in the top 2 cm be-

tween 4 and 251 ng Hg g

−1

dw and it is slightly surprising that not

even a trend of correlation is seen

–

indicating that the concentration

in the periwinkles is not dependent solely on the sediment concentra-

tion. A similar conclusion can be drawn from the investigation of mer-

cury in sediment and periwinkles of the contaminated Stour and Orwell

UK estuaries (Wright and Mason, 1999): mercury concentrations in

the upper 10 cm of the sediment at 29 sampling stations ranged from 60

P. Bjerregaard et al. / Environmental Pollution xxx (xxxx) 113985

5. Conclusion

Mercury concentrations in sediments and biota along Harboøre

Tange are considerably lower now than at the latest monitoring dur-

ing the 1980s with most sediment concentrations being below the Back-

ground Concentration (BC) of 50 ng Hg g

−1

dw defined by OSPAR. De-

spite the fact that some stations now have sediment concentrations at

the 2–10 ng Hg g

−1

dw level, characteristic of Danish coastal sediments

with no known history of mercury contamination, concentrations in

most samples of benthic biota along Harboøre Tange still exceed those in

uncontaminated coastal areas and also the Environmental Quality Stan-

dard (EQS) of 20 ng Hg g

−1

wet weight (~100 ng Hg g

−1

dry weight)

defined by the European Union's Water Framework Directive.

The slightly surprising observation that the mercury concentrations

in the benthic fauna exceed the EQS value despite sediment mercury

concentrations generally falling below the OSPAR BC and BAC thresh-

olds should be interpreted with care

–

especially in the light of our in-

complete understanding of the precise relation between mercury con-

centrations in sediment and organisms.

CRediT authorship contribution statement

Poul Bjerregaard:

Investigation, Writing - original draft.

Torben

Grau Schmidt:

Data curation, Formal analysis.

Maria Pedersen Mose:

Data curation, Formal analysis.

Declaration of competing interest

The authors declare no conflict of interests.

to 840 ng Hg g

−1

dw and average mercury concentrations in the peri-

winkles at 17 of the stations ranged between 60 and 990 ng Hg g

−1

and there is no correlation between the concentrations in sediment and

periwinkles (r = 0.33; p = 0.17 [calculated from data read from Figs.

2–4, 6 in Wright and Mason (1999)]). Also for the polychaete

Nereis

diversicolor

the data from Wright and Mason (1999) reveals no cor-

relation between the mercury concentrations in sediment and animal.

Contrary to this, Taylor et al. (2012) found clear correlations between

the total mercury (normalised with regard to organic content) in sedi-

ment and both polychates (Nereis sp.), gastropods (L.

littorea

and

Nas-

sarius obsoletus)

and shrimps (Crangon

septemspinosa

and

Palaemontes pu-

gio)

in Narragansett Bay with total mercury concentrations ranging be-

tween 35 and 2629 ng Hg g

−1

dw. Casado-Martinez et al. (2008)

found mercury concentrations in lugworms

Arenicola marina

from Span-

ish ports between 10 and 140 ng Hg g

−1

dw at sediment concentrations

between 50 and 32,000 ng Hg g

−1

dw without any clear correlation be-

tween the two. Correlation becomes apparent only at higher sediment

–

and lugworm

–

concentrations when expressed on a dry weight basis

(Casado-Martinez et al., 2008). However, if mercury concentrations

in the sediment are expressed on the basis of organic content, a correla-

tion (p = 0.002; data read from Casado-Martinez, Fig. 2b) becomes ap-

parent in the entire concentration range. Chen et al. (2009) found that

mercury concentrations in

C. maenas, M.edulis

and

L. littorea

varied no

more than a factor 2 to 4 among sites in the Gulf of Maine where sedi-

ment concentrations varied from 8 to 1135 ng Hg g

−1

dw and concluded

that sediment concentration is a poor predictor for concentrations in the

benthic biota.

Clear correlations between sediment mercury concentrations in the

Limfjord area and mercury concentrations in eelgrass leaves and roots

were identified by Lyngby and Brix (1987); mercury concentrations

in the leaves also showed a correlation with sediment mercury concen-

trations in the present investigation.

Acknowledgement

We thank Annette Duus, Anna-Katharina Semmler and Bente Frost

Holbech for assistance with the 2014-sampling. The investigation was

supported by a 50,000 D.Kr. grant (95-306-73551) from Lemvig Kom-

mune, Denmark, which holds the legal responsibility for the environ-

mental management of the area.

Appendix A. Supplementary data

Supplementary data to this article can be found online at

https://doi.

org/10.1016/j.envpol.2020.113985.

References

Akito, M., Shinichiro, Y., Akihiro, H., Michiaki, K., Ikuko, S., Akihide, T., Hirokatsu, A.,

2014. Reevaluation of Minamata Bay, 25 years after the dredging of mercury-polluted

sediments. Mar. Pollut. Bull. 89, 112–120.

Al-Zamel, A., Bou‐Rabee, F., Lesniewska, E., Szynkowska, M., Dlugosz, M., Bem, H., 2010.

Total mercury distribution in sediment cores from Kuwait Bay. Int. J. Environ. Stud.

67, 505–513.

Andersen, H.B., 1992. The expansion of mercury contamination, 5 years after discovery.

Mar. Pollut. Bull. 24, 367–369.

Azzaria, L.M., Aftabi, A., 1991. Stepwise thermal-analysis technique for estimating mer-

cury phases in soils and sediments. Water Air Soil Pollut. 56, 203–217.

Baeyens, W., Leermakers, M., De Gieter, M., Nguyen, H.L., Parmentier, K., Panutrakul, S.,

Elskens, M., 2005. Overview of trace metal contamination in the Scheldt estuary and

effect of regulatory measures. Hydrobiologia 540, 141–154.

Bakke, T.O., Kibsgaard, A., Breedveld, G., Eek, E., Helland, A., Källqvist, T., Ruus, A., Hyl-

land, K., 2007. In: Forurensningstilsyn (Ed.), Revidering Av Klassifisering Av Metaller

Og Organiske Miljøgifter I Vann Og Sedimenter. Norwegian.

Balogh, S.J., Tsui, M.T.K., Blum, J.D., Matsuyama, A., Woerndle, G.E., Yano, S., Tada, A.,

2015. Tracking the fate of mercury in the fish and bottom sediments of Minamata bay,

Japan, using stable mercury isotopes. Environ. Sci. Technol. 49, 5399–5406.

Beiras, R., Fernandez, N., Gonzalez, J.J., Besada, V., Schultze, F., 2002. Mercury concen-

trations in seawater, sediments and wild mussels from the coast of Galicia (NW Spain).

Mar. Pollut. Bull. 44, 345–349.

Beiras, R., Fernandez, N., Bellas, J., Besada, V., Gonzalez-Quijano, A., Nunes, T., 2003.

Integrative assessment of marine pollution in Galician estuaries using sediment chem-

istry, mussel bioaccumulation, and embryo-larval toxicity bioassays. Chemosphere 52,

1209–1224.

Beldowski, J., Miotk, M., Beldowska, M., Pempkowiak, J., 2014. Total, methyl and organic

mercury in sediments of the Southern Baltic Sea. Mar. Pollut. Bull. 87, 388–395.

Bouchet, S., Tessier, E., Monperrus, M., Bridou, R., Clavier, J., Thouzeau, G., Amouroux,

D., 2011. Measurements of gaseous mercury exchanges at the sediment-water, wa-

ter-atmosphere and sediment-atmosphere interfaces of a tidal environment (Arcachon

Bay, France). J. Environ. Monit. 13, 1351–1359.

Brix, H.L., Lyngby, J.E., 1984. In: Limfjordskomiteen (Ed.), Tungmetaller. Limfjorden

1982. 1. Kviksølv. Tungmetaller i Limfjorden, Danish, p. 139.

Casado-Martinez, M.C., Branco, V., Vale, C., Ferreira, A.M., DelValls, T.A., 2008. Is

Areni-

cola marina

a suitable test organism to evaluate the bioaccumulation potential of Hg,

PAHs and PCBs from dredged sediments? Chemosphere 70, 1756–1765.

Chen, C.Y., Dionne, M., Mayes, B.M., Ward, D.M., Sturup, S., Jackson, B.P., 2009. Mercury

bioavailability and bioaccumulation in estuarine food webs in the Gulf of Maine. Env-

iron. Sci. Technol. 43, 1804–1810.

Conaway, C.H., Squire, S., Mason, R.P., Flegal, A.R., 2003. Mercury speciation in the san

Francisco bay estuary. Mar. Chem. 80, 199–225.

Covelli, S., Faganeli, J., De Vittor, C., Predonzani, S., Acquavita, A., Horvat, M., 2008. Ben-

thic fluxes of mercury species in a lagoon environment (Grado lagoon, northern Adri-

atic Sea, Italy). Appl. Geochem. 23, 529–546.

Driscoll, C.T., Mason, R.P., Chan, H.M., Jacob, D.J., Pirrone, N., 2013. Mercury as a global

pollutant: sources, pathways, and effects. Environ. Sci. Technol. 47, 4967–4983.

DTU-Aqua, 2017. Hydrodynamics of the Limfjord. (in Danish).

During, A., Rinklebe, J., Boehme, F., Wennrich, R., Staerk, H.-J., Mothes, S., Du Laing, G.,

Schulz, E., Neue, H.-U., 2009. Mercury volatilization from three floodplain soils at the

Central Elbe river, Germany. Soil Sediment Contam. 18, 429–444.

EC, 2008. In: Union, E. (Ed.), Directive 2008/56/EC of the European Parliament and the

Council of 17 June 2008 Establishing a Framework for Community Action in the Field

of Marine Environment Policy (Marine Strategy Framework Directive), 164. Official J.

Eur. Union L, p. 19.

Emili, A., Acquavita, A., Koron, N., Covelli, S., Faganeli, J., Horvat, M., Zizek, S., Fajon,

V., 2012. Benthic flux measurements of Hg species in a northern Adriatic lagoon envi-

ronment (Marano and Grado Lagoon, Italy). Estuar. Coast Shelf Sci. 113, 71–84.

EU, 2008. Directive 2008/105/EC of the European Parliament and of the Council of 16

December 2008 on Environmental Quality Standards in the Field of Water Policy,

Amending and Subsequently Repealing Council Directives 82/176/EEC, 83/513/EEC,

84/156/EEC, 84/491/EEC, 86/280/EEC and Amending Directive 2000/60. EC of the

European Parliament and of the Council.

Gagnon, C., Fisher, N.S., 1997. Bioavailability of sediment-bound methyl and inorganic

mercury to a marine bivalve. Environ. Sci. Technol. 31, 993–998.

P. Bjerregaard et al. / Environmental Pollution xxx (xxxx) 113985

Wright, P., Mason, C.F., 1999. Spatial and seasonal variation in heavy metals in the sedi-

ments and biota of two adjacent estuaries, the Orwell and the Stour, in eastern Eng-

land. Sci. Total Environ. 226, 139–156.

Yano, S., 2013. In-situ measurement of mercury transport in the sea water of Minamata

bay. In: Matsui, K., Kramadibrata, S. (Eds.), International Conference on Earth Science

and Technology Proceedings. pp. 448–456.

Grandjean, P., Weihe, P., White, R.F., Debes, F., Araki, S., Yokoyama, K., Murata, K.,

Sorensen, N., Dahl, R., Jorgensen, P.J., 1997. Cognitive deficit in 7-year-old children

with prenatal exposure to methylmercury. Neurotoxicol. Teratol. 19, 417–428.

Kiorboe, T., Mohlenberg, F., Riisgard, H.U., 1983. Mercury levels in fish, invertebrates and

sediment in a recently recorded polluted area (Nissum Broad, western Limfjord, Den-

mark). Mar. Pollut. Bull. 14, 21–24.

Kudo, A., Fujikawa, Y., Mitui, M., Sugahara, M., Tao, G., Zheng, J., Sasaki, T., Miyahara,

S., Muramatsu, T., 2000. History of mercury migration from Minamata bay to the Yat-

sushiro Sea. Water Sci. Technol. 42, 177–184.

Kumamoto Prefecture, 2010. In: Kumamoto Prefecture (Ed.), Mercury Monitoring Report

of Minamata Bay (In Japanese). (Kumamoto City).

Lindberg, S.E., Dong, W.J., Meyers, T., 2002. Transpiration of gaseous elemental mer-

cury through vegetation in a subtropical wetland in Florida. Atmos. Environ. 36,

5207–5219.

Lindberg, S.E., Dong, W.J., Chanton, J., Qualls, R.G., Meyers, T., 2005. A mechanism for

bimodal emission of gaseous mercury from aquatic macrophytes. Atmos. Environ. 39,

1289–1301.

Long, E.R., Macdonald, D.D., Smith, S.L., Calder, F.D., 1995. Incidence of adverse biolog-

ical effects within ranges of chemical concentrations in marine and estuarine sedi-

ments. Environ. Manag. 19, 81–97.

Long, E.R., Field, L.J., MacDonald, D.D., 1998. Predicting toxicity in marine sediments

with numerical sediment quality guidelines. Environ. Toxicol. Chem. 17, 714–727.

Lyngby, J.E., Brix, H., 1987. Monitoring of mercury and cadmium in coastal areas, using

aquatic organisms and sediment. Water Sci. Technol. 19, 1239–1241.

Matsuyama, A., Yano, S., Hisano, A., Kindaichi, M., Sonoda, I., Tada, A., Akagi, H., 2016.

Distribution and characteristics of methylmercury in surface sediment in Minamata

Bay. Mar. Pollut. Bull. 109, 378–385.

Mohlenberg, F., Riisgard, H.U., 1988. Partitioning of inorganic and organic mercury in

cockles

Cardium edule

(L) and

Cardium glaucum

(Bruguiere) from a chronically pol-

luted area - influence of size and age. Environ. Pollut. 55, 137–148.

Monperrus, M., Tessier, E., Amouroux, D., Leynaert, A., Huonnic, P., Donard, O.F.X., 2007.

Mercury methylation, demethylation and reduction rates in coastal and marine sur-

face waters of the Mediterranean Sea. Mar. Chem. 107, 49–63.

OSPAR, 2009. In: OSPAR (Ed.), Background Document on CEMP Assessment Criteria for

the QSR 2010. p. 23.

Poissant, L., Constant, P., Pilote, M., Canario, J., O’Driscoll, N., Ridal, J., Lean, D.,

2007. The ebullition of hydrogen, carbon monoxide, methane, carbon dioxide and

total gaseous mercury from the Cornwall Area of Concern. Sci. Total Environ. 381,

256–262.

Rajar, R., Zagar, D., Cetina, M., Akagi, H., Yano, S., Tomiyasu, T., Horvat, M., 2004. Appli-

cation of three-dimensional mercury cycling model to coastal seas. Ecol. Model. 171,

139–155.

Ridgway, J., Shimmield, G., 2002. Estuaries as repositories of historical contamination and

their impact on shelf seas. Estuar. Coast Shelf Sci. 55, 903–928.

Riisgard, H.U., 1984. Mercury pollution in the waters around Harboore Tange and Lim-

fjord, Denmark. Mar. Pollut. Bull. 15, 129–133.

Riisgard, H.U., Famme, P., 1986. Accumulation of inorganic and organic mercury in

shrimp,

Crangon crangon.

Mar. Pollut. Bull. 17, 255–257.

Rinklebe, J., During, A., Overesch, M., Wennrich, R., Staerk, H.-J., Mothes, S., Neue, H.-U.,

2009. Optimization of a simple field method to determine mercury volatilization from

soils-Examples of 13 sites in floodplain ecosystems at the Elbe River (Germany). Ecol.

Eng. 35, 319–328.

Rinklebe, J., During, A., Overesch, M., Du Laing, G., Wennrich, R., Staerk, H.-J., Mothes,

S., 2010. Dynamics of mercury fluxes and their controlling factors in large Hg-polluted

floodplain areas. Environ. Pollut. 158, 308–318.

Rinklebe, J., Wennrich, R., Du Laing, G., Staerk, H.J., Mothes, S., 2013. Mercury emissions

from flooded soils and sediments in Germany are an underestimated problem: chal-

lenges for reliable risk assessments and management strategies. In: Pirrone, N. (Ed.),

Proceedings of the 16th International Conference on Heavy Metals in the Environ-

ment.

Scheuhammer, A.M., Basu, N., Burgess, N.M., Elliott, J.E., Campbell, G.D., Wayland, M.,

Champoux, L., Rodrigue, J., 2008. Relationships among mercury, selenium, and neu-

rochemical parameters in common loons (Gavia immer) and bald eagles (Haliaeetus

leucocephalus). Ecotoxicology 17, 93–101.

Schultz, V., 2017. The Fate of Mercury in Danish Coastal Systems: May Divalent Merury Be

Reduced to Elementary Mercury I Sediments? (In Danish) Master thesis. Department

of Biology, University of Southern Denmark, p. 52.

Sharif, A., Tessier, E., Bouchet, S., Monperrus, M., Pinaly, H., Amouroux, D., 2013. Com-

parison of different air-water gas exchange models to determine gaseous mercury eva-

sion from different European coastal lagoons and estuaries. Water Air Soil Pollut. 224.

Spada, L., Annicchiarico, C., Cardellicchio, N., Giandomenico, S., Di Leo, A., 2012. Mer-

cury and methylmercury concentrations in Mediterranean seafood and surface sed-

iments, intake evaluation and risk for consumers. Int. J. Hyg Environ. Health 215,

418–426.

Taylor, D.L., Linehan, J.C., Murray, D.W., Prell, W.L., 2012. Indicators of sediment and bi-

otic mercury contamination in a southern New England estuary. Mar. Pollut. Bull. 64,

807–819.

Tomiyasu, T., Matsuyama, A., Eguchi, T., Fuchigami, Y., Oki, K., Horvat, M., Rajar, R., Ak-

agi, H., 2006. Spatial variations of mercury in sediment of Minamata Bay, Japan. Sci.

Total Environ. 368, 283–290.

Tomiyasu, T., Takenaka, S., Noguchi, Y., Kodamatani, H., Matsuyamab, A., Oki, K., Kono,

Y., Kanzaki, R., Akagi, H., 2014. Estimation of the residual total mercury in marine

sediments of Minamata Bay after a pollution prevention project. Mar. Chem. 159,

19–24.

UNEP, 2013. Global Mercury Assessment 2013. Sources, Emissions, Releases and Environ-

mental Transport. United Nations Environment Programme, Geneva, p. 36.

Walker, T.R., 2016. Mercury concentrations in marine sediments near a former mercury

cell chlor-alkali plant in eastern Canada. Mar. Pollut. Bull. 107, 398–401.