ARTICLE IN PRESS

Environmental Research 109 (2009) 355–367

Contents lists available at ScienceDirect

Environmental Research

0013-93

doi:10.1

� Corr

E-m

journal homepage: www.elsevier.com/locate/envres

Human exposure to mercury in the vicinity of chlor-alkali plant

Darija Gibicar a, Milena Horvat a,�, Martina Logar a, Vesna Fajon a, Ingrid Falnoga a, Romano Ferrara b,Enrica Lanzillotta b, Claudia Ceccarini b, Barbara Mazzolai c, Bruce Denby d, Jozef Pacyna d,e

a Jozef Stefan Institute, Department of Environmental Sciences, Jamova cesta 39, 1000 Ljubljana, Sloveniab Institute of Biofisics Area della Ricerca, Pisa, Italyc Scuola Superiore Sant’Anna, CRIM Laboratory, Pisa, Italyd Norwegian Institute for Air Research, PO Box 100, No-2027 Kjeller, Norwaye Faculty of Chemistry, Gdansk University of Technology, 11/12 G. Narutowicza Str., 80-952 Gdansk, Poland

a r t i c l e i n f o

Article history:

Received 23 July 2008

Received in revised form

1 January 2009

Accepted 23 January 2009Available online 14 March 2009

Keywords:

Chlor-alkali plant

Human exposure

Mercury

Methylmercury

Selenium

Vegetable

Fish

Air

51/$ - see front matter & 2009 Elsevier Inc. A

016/j.envres.2009.01.008

esponding author. Fax:+386 15885346.

ail address: milena.horvat@ijs.si (M. Horvat).

a b s t r a c t

The main objectives of our study were to estimate the impact of a mercury cell chlor-alkali (MCCA)

complex in Rosignano Solvay (Tuscany, Italy) on the local environment and to assess mercury exposure

of inhabitants living near the plant. Measurement campaigns of atmospheric Hg near the MCCA plant

showed that the impact of the emitted Hg from the industry on the terrestrial environment is restricted

to a close surrounding area. Total gaseous mercury concentrations in ambient air of inhabited area

around the MCCA plant were in the range of 8.0–8.7 ng/m3 in summer and 2.8–4.2 ng/m3 in winter.

Peaks of up to 100 ng/m3 were observed at particular meteorological conditions. Background levels of

2 ng/m3 were reached within a radius of 3 km from the plant. Reactive gaseous mercury emissions from

the plant constituted around 4.2% of total gaseous mercury and total particulate mercury emission

constituted around 1.0% of total gaseous mercury emitted. Analysis of local vegetables and soil samples

showed relatively low concentrations of total mercury (30.1–2919mg Hg/kg DW in the soil; o0.05–111

mg Hg/kg DW in vegetables) and methylmercury (0.02–3.88mg Hg/kg DW in the soil; 0.03–1.18mg Hg/kg

DW in vegetables). Locally caught marine fish and fresh marine fish from the local market had

concentrations of total Hg from 0.049 to 2.48mg Hg/g FW, of which 37–100% were in the form of

methylmercury. 19% of analysed fish exceeded 1.0mg Hg/g FW level, which is a limit set by the European

Union law on Hg concentrations in edible marine species for tuna, swordfish and shark, while 39% of

analysed fish exceeded the limit of 0.5mg Hg/g FW set for all other edible marine species. Risk

assessment performed by calculating ratio of probable daily intake (PDI) and provisional tolerable daily

intake (PTDI) for mercury species for various exposure pathways showed no risks to human health for

elemental and inorganic mercury, except for some individuals with higher number of amalgam fillings,

while PDI/PTDI ratio for methylmercury and total mercury exceeded the toxicologically tolerable value

due to the potential consumption of contaminated marine fish.

& 2009 Elsevier Inc. All rights reserved.

0. Introduction

Mercury is a naturally occurring element in the Earth’s crust.Over geological time, it has been distributed throughout theenvironment by natural processes, such as volcanic activity, fires,movement of rivers, lakes, and streams, oceanic up-welling, andbiological processes. Since the advent of humans, and particularlysince the industrial revolution of the late 18th and 19th centuries,anthropogenic sources have become a significant contributor tothe environmental distribution of mercury and its compounds(WHO, 2003). Major human sources of mercury involve coalcombustion in power plants, the production of caustic soda withthe use of the Hg cell process, and cement production (Pacyna

ll rights reserved.

et al., 2006). Although the mercury cell chlor-alkali (MCCA)technique is in many places being replaced by alternativetechniques, it is still the most commonly used in Europe. Morerecent estimates report emissions from the chlor-alkali industry tobe responsible for about 17% or 40.4 ton/yr of anthropogenic totalmercury emissions (Pacyna et al., 2006).

In Tuscany, Italy, mercury is present in the environment as aproduct of natural and anthropogenic sources. The cinnabar (HgS)deposits of Mt. Amiata are a major source of Hg in the central partof Tuscany. Mercury mining and smelting activities in this areawere present as early as the Etruscan period (8th–1st centuriesB.C.) and ceased in 1980. In this area high mercury levels werefound in soil, vegetation, air (Bargagli et al., 1987; Barghigiani andBauleo, 1992; Ferrara et al., 1992), farm produce (Barghigiani andRistori, 1994) and fish (Barghigiani et al., 1991; Barghigiani and DeRanieri, 1992; Rossi et al., 1993; Barghigiani et al., 2000; Scerboet al., 2005). Additional sources of mercury contamination in this

ARTICLE IN PRESS

D. Gibicar et al. / Environmental Research 109 (2009) 355–367356

region are geothermal power plants (Bargagli and Barghigiani,1991) that mobilise natural mercury in the central part of theregion. Moreover two MCCA plants are located in Tuscany causingadditional mercury pollution. One is Europe’s largest MCCA plantlocated on the coast near Livorno in Rosignano Solvay thatdischarges waste directly in the Tyrrhenian Sea, and the other oneis located in a small village Saline di Volterra, which dischargeswaste into the S. Marta, a canal flowing into the river Cecina(Scerbo et al., 2005).

The chlor-alkali complex in Rosignano Solvay has been inoperation since 1940. Its impact on the surrounding environmenthas been studied by several investigators since 1970 (Renzoniet al., 1973; Baldi and Bargagli, 1984; Baldi and D’Amato, 1986;Ferrara et al., 1989, 1992, 2001; Maserti and Ferrara, 1991). Masertiand Ferrara (1991) reported that before 1973, a yearly discharge ofHg from the MCCA plant into the coastal seawater was between12,500 and 14,500 kg. In 1975–1976, after the waste treatmentfacilities were installed (Baldi and D’Amato, 1986), the reporteddischarge was 350–750 kg Hg/yr. In 2001 the discharge of Hg wasestimated at 400–500 kg Hg/yr (Ferrara et al., 2001). Oncedischarged Hg enters in the aquatic ecosystems, part of theinorganic mercury (I-Hg) can be microbially converted intomethylmercury (MeHg) and taken up by aquatic organisms. Fishaccumulate mercury directly from food and the surroundingwater (Rainbow, 1985). They can bioconcentrate large amounts ofthis metal and seafood is considered the single largest source ofmethylmercury in humans (Clarkson, 1997).

The main objectives of our study were to assess the impact ofmercury emission from the MCCA plant in Rosignano Solvay onthe local environment. We determined Hg species in the atmo-sphere of the village surrounding the MCCA plant, the depositionof Hg from the plant, levels of total mercury (THg) andmethylmercury in the soil and in the locally grown vegetables.The results were compared with the levels of Hg from the ruralreference area, situated 20 km south from the plant (villageDonoratico). Additionally we determined levels of total andmethylmercury in various fish caught from the Rosignano Solvayarea, Castiglion della Pescaia and vicinity of Elba Island and fishbought at the local market. Finally we performed a robust exposureassessment for inhabitants living near the MCCA plant andcompared it with the exposure for inhabitants of the reference area.In order to evaluate whether there is an increased risk of adversehealth effects from elemental, inorganic, methylmercury and totalmercury exposure, simple risk assessment was performed byestimating probable daily intakes (PDI) of elemental Hg, inorganic,methylmercury and total mercury and comparing them with theprovisional tolerably daily intakes (PTDI) of mercury.

The study was carried out in the years 2002–2004 within theEuropean Union funded project named EMECAP (EuropeanMercury Emissions from Chlor Alkali Plants).

1. Methods

1.1. Study location

The village Rosignano Solvay is situated in Tuscany along the Tyrrhenian coast,

500 m from the shore-line, about 20 km south of Livorno (Fig. 1). The plant has

been in operation since 1940 (Maserti and Ferrara, 1991). The chlorine production

in the Solvay chlor-alkali plant nowadays is around 120,000 ton/yr (Euro Chlor,

2005). The village Donoratico, located 20 km south of Rosignano Solvay, was

selected as the most suitable reference area. It has no industrial plants in the

immediate vicinity with a similar life style of inhabitants as in Rosignano.

1.2. Atmospheric mercury and modelling of Hg deposition

Four seasonal measurement campaigns were carried out during winter and

summer in years 2002 and 2003 in order to determine the following parameters:

total gaseous mercury (TGM) concentration, reactive gaseous mercury (RGM)

concentration, total particulate mercury (TPM) concentration, Hg flux from soil

and Hg content in rain. Nine sampling stations were selected in the village of

Rosignano Solvay around the chlor-alkali plant, taking into account where citizens

live, and 1 sampling station in the reference area.

Total gaseous mercury (TGM) refers to gaseous elemental mercury (GEM) and

small contributions from other gaseous mercury species (less than 1%) that also

may be trapped by the sampler and detected as Hg0. Daily trend of TGM

concentrations has been determined by an automated Gardis-3 analyser

(Ekoservis, Lithuania) using the double amalgamation technique followed by cold

vapour atomic absorption spectrometry (CV AAS). Detailed procedure and

comparison of results in measurements between GARDIS, TEKRAN and the LIDAR

set-up were described by Wangberg et al. (2003), confirming a high degree of

analytical quality for mercury measurements in air.

RGM is operationally defined gaseous mercury fraction present in ambient air.

It is believed that RGM mostly consists of mercury dichloride (HgCl2), but other

divalent mercury species are also conceivable. Mist chamber, containing a KCl/HCl

solution has been used to trap RGM. The solution was then analysed by SnCl2

reduction and CVAFS (Wangberg et al., 2003).

Daily average concentrations of TPM at two stations have been determined by

quartz fibre filter traps. The details of the method have been described previously

by Wangberg et al. (2003).

The deposition of mercury near the source has been evaluated on the base of

mercury concentration in rain samples. Hg flux from soil has been determined by

flux chamber technique, a detailed procedure was described elsewhere (Xiao et al.,

1991; Kim and Lindberg, 1995; Wangberg et al., 2003).

‘The Atmospheric Pollution Model’ (TAPM) from CSIRO in Australia was used

for meteorological calculations and an off-line dispersion chemistry model

EPISODE has been adapted to include a mercury–chlorine chemistry scheme in

modelling process and to calculate deposition and concentration fields of GEM,

RGM and TPM. A mercury/chlorine/ozone chemistry scheme was developed to

describe chemistry in the plume as well as in the factory itself. For the local scale

modelling the development and testing of a suitable dispersion modelling system

to describe local scale mercury dispersion and chemistry was performed. The

model was validated against observed concentration and emission data, collected

during the measurement campaigns. The dispersion and deposition of mercury

emitted from the MCCA plant was simulated for the year 2002 and local

concentration and deposition fields were calculated. Detailed modelling descrip-

tion was described elsewhere (Denby and Pacyna, 2004).

1.3. Sampling of soil, vegetables and fish

Soil samples were collected during October of 2001 and 2002 and July 2002

from 15 gardens in the Rosignano Solvay area (Fig. 2) and the Donoratico area as

the control station (4 sampling points). Surface soil samples were transferred to

plastic bags and stored in a freezer. Homogenisation was done by gentle mixing

using a glass stick. Dry weight was determined in separate aliquots by heating to

constant weight at 105 1C.

From the same sampling points vegetable samples (celery, salad, fennel,

tomato, beet, basil, radish, parsley, aubergines, garlic, onion, bean, celery, gourd,

potato) were collected during October of 2001 and 2002 and July 2002. Collected

vegetable samples were rinsed with tap water, air-dried and transferred to plastic

bags and stored deep-frozen. Prior to homogenisation in agate mortar vegetable

samples were freeze-dried. Before and after freeze-drying the weight was recorded

in order to be able to express the data on both fresh (as consumed) and dry weight

bases.

Nineteen fish and 1 crab were collected offshore Rosignano Solvay, Castiglion

della Pescaia and the Island of Elba and 6 fish were bought in the Piombino

market. Fish were collected in October 2001 and April 2002. Sampling locations

are shown in Fig. 1 and Table 4. Samples belonged to species Trachinus araneus

(striped weever), Trigla lucerna (tub gurnard), Scorpaena notata (scorpion fish),

Serranus scriba (painted comber, perch), Uranoscopus scaber (stragazer),

Conger conger (conger eel), Lichia amia (leer fish, jack), Pomatomus saltatrix

(bluefish) and Eriphia verrucosa (crab). Samples were stored in a freezer below

20 1C. Prior the analysis fish samples were slightly thawed, the skin removed and

fish muscle carefully removed and homogenized. Fresh muscle samples were

analysed.

1.4. Analytical methods for determination of Hg species and total selenium in air, soil,

vegetables and fish

Various analytical methods were used which are described elsewhere.

A summary of the analytical methods is presented in Table 1. The accuracy of

the results was checked by the use of certified reference materials for total and

methylmercury determination for each batch of the samples. We used NIST SRM

2976 (Mussel homogenate, assigned values: 61.073.5 ng THg/g, 27.772.0 ng

MeHg/g), NIST SRM 2977 (Mussel tissue, assigned value: 97–105 ng THg/g), NRCC

DOLT-2 (Dogfish liver, assigned values: 199071000 ng THg/g, 693753 ng MeHg/g)

ARTICLE IN PRESS

Fig. 1. Map of Tuscany (Italy) with marked sampling locations: the Rosignano Solvay mercury cell chlor-alkali plant in red square, reference station (Donoratico) in yellow

triangle, blue circles indicating origin of analysed fish. The distance between Riosignanio and Donoratico is about 20 km. For interpretation of the references to colour in this

figure legend, the reader is referred to the web version of this article.

D. Gibicar et al. / Environmental Research 109 (2009) 355–367 357

and NRCC TORT 2 (Lobster Hepatopancreas, assigned value: 270760 ng THg/g) to

control the analytical accuracy for Hg determination in fish; IAEA 405 (Estuarine

sediment, assigned values: 770–850 ng THg/g, 5.4970.05 ng MeHg/g) to control

the analytical accuracy for Hg determination in soil samples, IAEA 336 (Lichen,

assigned value: 167–233 ng THg/g) and IAEA 140 (Fucus; sea plant homogenate,

assigned values: 32–44 ng THg/g, 0.62670.107 ng MeHg/g) to control the analy-

tical accuracy for Hg determination in vegetable samples. The mean levels of total

and methylmercury in NIST 2976 (63.175.1 ng THg/g, N ¼ 25; 27.072.0 ng

MeHg/g, N ¼ 4), NIST 2977 (10377 ng THg/g, N ¼ 5), DOLT-2 (18167160 ng THg/g,

N ¼ 3; 650718 ng MeHg/g, N ¼ 3), TORT 2 (273 ng THg/g, N ¼ 1), IAEA 405

(800711 ng THg/g, N ¼ 3; 5.2270.10 ng MeHg/g, N ¼ 5), IAEA 336 (19175 ng

THg/g, N ¼ 3), IAEA 140 (33.470.6 ng THg/g, N ¼ 3, 0.6070.05 ng MeHg/g,

N ¼ 3), were well within their range given for the assigned values.

1.5. Exposure assessment

To determine possible exposure to mercury by various pathways, such as

inhalation and consumption of contaminated food, probable daily intake values

were calculated for the general adult population. The equation used for calculating

total daily exposure dose was

PDItotal ¼ PDIinhalation þ PDIingestion

PDIingestionðmg=kg body weight=dayÞ ¼ food intakeðg=dayÞ

� ½I- or MeHg concentrationðmg=gÞ�average body weightðkgÞ

PDIinhalationðmg=kg body weight=dayÞ ¼ daily respiratory volume ðm3Þ

� ½Hg0 concentrationðng=m3Þ� � 10�3 average body weight ðkgÞ

Average daily fish intake was obtained from the European Food Safety Authority

(EFSA, 2004), where 32 g/day of fish for Italy is reported. Average body weight for

an adult Italian is estimated at 70 kg (EFSA, 2004). We estimated an average daily

vegetable intake to be around 300 g. Measured concentrations of mercury in air,

fish and vegetables as sources of Hg were considered and taken from Table 2. For

Hg in drinking water, amalgam fillings and food other than fish and vegetables,

average daily Hg intakes reported by WHO (2003) were included. Daily respiratory

volume of 20 m3 was considered.

Estimated PDI values were considered against the provisional tolerable daily

intake (PTDI) to evaluate whether there is an increased risk of adverse health

effects from elemental Hg, inorganic Hg, methylmercury and total mercury

exposure. A comparison of the two parameters was achieved by determining the

ratio of the PDI to the PTDI, expressed as a percentage (PDI/PTDI�100%). Values

approaching or exceeding 100% identify those exposure scenarios where the

toxicological reference value may be exceeded and that require more careful

evaluation.

1.6. Statistical analysis

Standard descriptive statistics were produced for the various variables and

strata, with use of the medians and/or geometric means for most variables.

STATGRAPHIC Plus for Windows Version 4.0 and SPSS 12.0.1 for Windows were

used as software. One-way ANOVA for sample comparison was used and

Mann–Whitney’s test and Kruskal–Wallis’ test at 95.0% confidence level for post

hoc testing. To investigate associations between the attributes, Spearman’s rank

correlations and Pearson’s correlations on log transformed data were used.

2. Results

2.1. Levels of atmospheric mercury in Rosignano Solvay and

modelling of its distribution and deposition

TGM concentrations in Rosignano Solvay were in the range of8.0–8.7 ng/m3 in summer and 2.8–4.2 ng/m3 in winter, althoughpeaks of up to 100 ng/m3 were observed in particular meteor-ological conditions (Table 2). At the reference station TGMconcentrations ranged between 1.5 and 5.5 ng/m3 (Table 2).Reactive gaseous mercury (HgII) concentrations measured inRosignano Solvay, most likely mercury chloride homologues(277–691 pg/m3), were higher compared to the reference stationin Donoratico (87–326 pg/m3).

Concentrations of mercury associated with particulate matter(TPM) ranged from 25 to 70 pg/m3 in Rosignano Solvay while inthe control station the range was 13–26 pg/m3. Rain samples from

ARTICLE IN PRESS

Fig. 2. Sampling stations around the mercury cell chlor-alkali plant in Rosignano Solvay. The position of the chlor-alkali plant is circled with the red line. Red square

indicates the cell of the plant. For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.

D. Gibicar et al. / Environmental Research 109 (2009) 355–367358

Rosignano Solvay contained 6–15 ng Hg/L and from Donoratico4.6–5.5 ng Hg/L. Measurements of mercury degassing rate fromsoil around the chlor-alkali plant was estimated to be 0–45 ng/m2/h.Mercury degassing rate from soil in reference station was lower andranged 0–2.5 ng/m2/h (Table 2).

The dispersion and deposition of mercury emitted from theMCCA plant were simulated for the year 2002. The results areshown in Fig. 3, where TGM concentrations and deposition ofTGM as a function of distance from the plant are shown.Heightened concentrations of TGM were found only in the regionclose to the plant and background levels of 2 ng/m3 were reachedwithin 3 km of the plant.

Emissions of mercury from the plant were estimated to bearound 285 kg/yr (Gronlund et al., 2005). Our results show thataround 4.2% of total emitted Hg was present in the form of HgII,and only around 1.0% in the form of TPM, while most of theemitted mercury was in its elemental form Hg0. Only 14% (or6.8 kg/yr) of total emitted Hg were deposited within 5 km fromthe plant. The remaining 86% were dispersed and transportedaway from the area, showing its characteristics of a globalpollutant. It is known that mercury may travel as far as 2500 kmin just 72 h and estimates of airborne residence time range from6 days to 6 years (US EPA, 2001), before the mercury isredeposited in air or water by rainfall or other climatologicalconditions.

2.2. Total and methylmercury in soil from Rosignano Solvay

Soil gathered from 15 gardens located around the MCCA plantin Rosignano Solvay showed relatively low concentrations of totaland methylmercury (Table 2). Concentrations of THg rangedbetween 30.1 and 2919mg/kg DW (Fig. 4) and were similar toconcentrations found by Maserti and Ferrara in 1991 on the samearea.

Average concentrations of THg and MeHg in soil from theRosignano Solvay area were slightly higher (but not statisticallysignificant at the 95.0% confidence level) than concentrationsfound in the reference area (Table 2). Total mercury in soil waselevated (2920mg/kg DW) only at one station of Rosignano Solvayarea, which was situated near the chlor-alkali plant in the westerndirection. None of the analysed soil samples exceeded the SoilGuideline Value for inorganic mercury contamination set byDepartment for Environment, Food and Rural Affairs (DEFRA,2002) (8 mg/kg DW) or Canadian Soil Quality Guideline foragricultural soils (6.6 mg/kg) or the UK soil guideline value forinorganic Hg for allotments (8 mg/kg). Mean concentrations ofTHg and MeHg were only slightly higher in soil samples collectedin autumn than those collected in summer, but the differencewas not statistically significant at 95% confidence level. MeanMeHg/THg ratio in soil collected in autumn (0.61%) was higherthan the mean MeHg/THg ratio in soil collected in summer

ARTICLE IN PRESS

Table 1Summary of analytical methods used for the analysis of total Hg, MeHg and selenium.

Sample, analyte Methods

Pretreatment/extraction/digestion Detection Limit of detection Reference

Air, Hg0 Amalgamation on gold trap AAS Gardis-3 0.1 ng/m3 Wangberg et al.

(2003)

Air, HgII Mist chamber-adsorption on KCl/HCl solution-reduction

with SnCl2-amalgamation on gold trap

AFS 5 pg/m3 Wangberg et al.

(2003)

Air, Hg(p) Quartz fibre filtre traps-reduction by pyrolysis AAS 5 pg/m3 Wangberg et al.

(2003)

Hg degassing rate

from soil

Flux chamber-amalgamation on gold trap AAS Gardis-3 0.1 ng/m3 Wangberg et al.

(2003)

Soil, total Hg Acid digestion with HNO3 and H2SO4 in closed Teflon vials-

reduction with SnCl2-amalgamation on gold trap

CV AAS 0.05 ng/g Horvat et al. (1991)

Vegetable, total Hg Acid digestion with HNO3 in closed Teflon vials-reduction

with SnCl2-amalgamation on gold trap

CV AAS 0.05 ng/g Horvat et al. (1991)

Fish, total Hg Acid digestion with HNO3 in closed Teflon vials-reduction

with SnCl2-amalgamation on gold trap

CV AAS 0.05 ng/g Horvat et al. (1991)

Vegetable, soil, MeHg Acidification with HCl-extraction of MeHgCl into CH2Cl2-

back extraction of MeHgCl into water phase by evaporation of

CH2Cl2-ethylation of ionic Hg species-room temperature

adsorption on Tenax

Isothermal gas

chromatography,

pyrolysis, CV AFS

0.01 ng/g Horvat et al. (1993)

Liang et al.

(1994a, b)

Fish, MeHg Acid leaching with KBr/H2SO4-extraction of MeHgBr into

toluene- clean-up with aqueous solution of cysteine-

re-extraction of MeHg into benzene

GC ECD 0.5 ng/g Horvat et al. (1988,

1990, 1994)

Fish, HgII and MeHg Alkaline dissolution with KOH/methanol-aqueous phase

ethylation room temperature adsorption on Tenax

CV AFS 30 pg/g for MeHg and

80 pg/g for Hg2+

Logar et al. (1999)

Liang et al. (1994b)

Vegetable, Se Irradiation in TRIGA Mark II reactor at a flux of 2�1010 n/mm2/s NAA, Gamma

spectrometry

10 ng/g Byrne and

Kosta (1974)

Table 2Summary statistics for concentrations of TGM, RGM and TPM in air, and for concentrations of THg, MeHg (expressed as Hg) and MeHg/THg ratios and Se concentrations

in soil, vegetables and rain samples from Rosignano Solvay and the reference area.

Rosignano solvay Reference area (Donoratico)

Mean7Std. dev. (n) Median Min–Max Mean7Std. dev. (n) Median Min–Max

Air

TGM (ng/m3)

Winter 2.8–4.2 1.5–5.5

Summer 8.0–8.7 (100)

RGM (pg/m3) 277–691 87–326

TPM (pg/m3) 25–70 13–26

Rain

THg (ng/L) 6–15 4.6–5.5

Soil

THg (mg/kg DW) 2047480 (36) 84.5 30.1–2919 1257113 (6) 112 30.4–341

MeHg (mg/kg DW) 0.6670.68 (36) 0.43 0.02–3.88 0.3670.32 (6) 0.34 0.03–0.88

MeHg/THg (%) 0.6370.45 (36) 0.51 0.01–1.76 0.6370.95 (6) 0.28 0.01–2.54

Degassing rate (ng Hg/m2/h) 0–45 0–2.5

Vegetables

THg

(mg/kg DW) 19.5720.5 (94) 16.2 o0.05–111 13.2713.2 (21) 9.92 0.22–54.7

(mg/kg FW) 3.4674.78 (94) 1.45 o0.01–24.0 3.0073.94 (21) 2.12 0.03–17.1

MeHg

(mg/kg DW) 0.2770.21 (94) 0.20 0.03–1.18 0.2170.15 (21) 0.15 0.09–0.72

(mg/kg FW) 0.0570.06 (94) 0.02 0.002–0.348 0.0570.06 (21) 0.02 0.01–0.29

MeHg /THg (%) 10.8721.1 (94) 1.87 0.09–87.0 7.10712.7 (21) 1.63 0.31–53.5

Se

(mg/kg DW) 17.9713.4 11.6 o10–58.2 14.4711.5 (21) 8.80 o10–47.6

(mg/kg FW) 4.374.3 2.7 o0.5–22.8 2.3472.13 (21) 1.86 0.00–9.98

D. Gibicar et al. / Environmental Research 109 (2009) 355–367 359

ARTICLE IN PRESS

Fig. 3. Concentration fields of: (a) average annual Hg0 concentrations in air, expressed in ng/m3, (b) average mercury deposition rates, expressed in ng/m2/h. Concentration

fields were calculated for the year 2002. The red square shows the position of the cell of the MCCA plant. For interpretation of the references to colour in this figure legend,

the reader is referred to the web version of this article.

D. Gibicar et al. / Environmental Research 109 (2009) 355–367360

(0.25%). MeHg/THg ratios in soil are probably related primarily tothe combined soil environmental conditions such as soil organicmatter nature, microbial activity, pH and THg concentration thataffect demethylation (Remy et al., 2006). The MeHg/THg ratio insoil samples from Rosignano Solvay was relatively low (median0.51%) and was slightly higher (but not statistically significant at95% confidence level) than the MeHg/THg ratio in soil from thereference area (median 0.28%) (Table 2).

2.3. Total mercury, methylmercury and selenium content in

vegetables from Rosignano Solvay and Donoratico area

We analysed vegetables grown in 15 gardens located aroundthe MCCA plant and compared them to concentrations invegetables grown at the reference site. In general vegetablescontained relatively low concentrations of total and methylmer-cury (Tables 2 and 3). Cabbage, salad, tomato, beet and basil fromthe Rosignano Solvay area showed slightly higher concentrationsof THg and MeHg compared to the reference area, but thedifference was not statistically significant at the 95% confidencelevel. Mean THg and MeHg concentrations in analysed vegetablesdiffered significantly between the vegetable species (THg:F ¼ 4.36, p ¼ 0.0000; MeHg: F ¼ 2.43, p ¼ 0.0038). The highestTHg concentrations expressed on a dry weight basis were found inbasil, beet, celery, parsley and mint (Table 3).

The MeHg/THg ratio in vegetables varied statistically signifi-cant at the 95% confidence level between the vegetable species(F ¼ 5.24, p ¼ 0.0000) and between the sampling locationsaround the MCCA plant (F ¼ 1.82, p ¼ 0.0327), but not betweenthe Rosignano Solvay area and the reference area. The highestMeHg/THg ratio was observed in aubergine, bean, garlic, gourd,onion and tomato. In general, taking into account all vegetablespecies, a positive Spearman’s rank correlation was foundbetween MeHg and THg (R2

¼ 0.4368, p ¼ 0.0001) and betweenTHg and MeHg/THg ratio (R2

¼ �0.7826, p ¼ 0.0000).

Mercury and selenium co-accumulation in 1:1 formation inhumans could be an important Hg detoxification process (Kostaet al., 1975; Byrne et al., 1995; Falnoga et al., 2000) and the lack ofselenium in the body can deteriorate the antioxidative role ofselenoproteins, such as glutathione peroxidase, formate dehydro-genase and selenophosphate synthase. The Se deficiency inhuman diet is more common than its toxicity, because manyregions have low levels of soil selenium. Some countriessupplement agricultural fertilizers with selenium. Vegetablesfrom Rosignano Solvay and the reference area contained relativelylow levels of selenium, which ranged from o10 to 58.2mg/kg DWor o0.5 to 22.8mg/kg FW. No statistically significant differencesbetween the sampling stations or between the species werefound. The highest selenium concentrations were found in basil,onion and salad (Table 3). Selenium in vegetables was notassociated with the content of THg or MeHg/THg ratio in theplants. However weak positive association between log(Se) andlog(MeHg) was observed (R2

¼ 0.2229, p ¼ 0.0484).Assuming daily consumption of 300 g of locally grown

vegetable, the daily intake of selenium would range from 0.15 to6.84mg. Intake of selenium by vegetables in this region is poor.The recommendations of US National Academy of Sciences setdaily intake of selenium in the range of 50–200mg as safe andadequate (McLaughlin et al., 1999; Reilly, 1996).

2.4. Mercury concentrations in local fish

The concentrations of total Hg in 26 analysed fish samplesvaried from 0.049 to 2.48 mg/kg of which 37–100% were found inthe form of methylmercury (Table 4). The highest mercury contentwas found in fish from the local market in Piombino. Five of the26 analysed samples (or 19%) exceeded 1.0 mg/kg FW level, whichis a limit set by the European Union law on Hg concentrations inedible marine species for tuna, swordfish and shark. Ten of the26 analysed samples (or 39%) exceeded the limit set for all other

ARTICLE IN PRESS

Fig. 4. Mean mercury concentrations of total mercury (mg/kg DW) in soil samples from 15 sampling points around the MCCA plant in Rosignano Solvay and from the

reference area (village Donoratico).

Table 3Mercury, methylmercury (expressed as Hg) and selenium in vegetables from Rosignano Solvay and Donoratico, expressed as mg/kg of dry weight and fresh weight.

Vegetable THg MeHg MeHg/THg Se

(mg/kg DW) (mg/kg FW) (mg/kg DW) (mg/kg FW) (mg/kg DW) (mg/kg FW)

Cabbage

R. Solv. (8)a 24.2710.9b 3.872.2 0.1570.13 0.0270.01 0.670.4 13.679.9 2.771.6

Donoratico (3) 12.27 8.3 2.171.3 0.1670.09 0.0370.01 1.670.8 12.074.8 2.371.2

Salad

R. Solv. (15) 26.5717.6 2.271.9 0.2570.12 0.0270.02 2.072.4 20.4710.9 2.872.4

Donoratico (3) 10.874.6 1.170.6 0.1570.01 0.0170.00 1.570.7 23.6721.3 2.572.3

Fennel

R. Solv. (9) 18.1716.0 1.271.0 0.3370.18 0.0270.01 2.772.0 – –

Donoratico (0) – – – – – – –

Tomato

R. Solv. (17) 3.878.8 0.470.8 0.1270.07 0.0170.01 14.6721.3 12.076.7 1.270.6

Donoratico (4) 0.4470.27 0.0470.02 0.1070.02 0.0170.00 21.1723.0 11.475.5 1.070.5

Beet

R. Solv. (8) 44.2735.2 3.774.2 0.3170.17 0.0270.02 2.875.7 o10 udl

Donoratico (5) 17.776.6 2.671.1 0.1770.05 0.0370.01 1.170.5 8.473.2 1.270.5

Basil

R. Solv. (12) 32.5717.7 9.875.5 0.4070.28 0.1370.09 1.470.8 23.5718.6 7.376.7

Donoratico (4) 24.3721.4 6.976.9 0.3070.11 0.0970.04 2.071.4 12.077.9 3.171.0

Mint

R. Solv. (1) 47.9 24.0 0.23 0.11 0.5 11.1 5.6

Donoratico (0) – – – – – – –

Parsley

R. Solv. (5) 26.376.7 9.872.9 0.3470.34 0.1270.10 1.371.2 13.674.7 5.774.0

Donoratico (2) 28.0717.1 9.374.2 0.6170.16 0.2270.10 2.972.3 21.9714.9 7.373.9

Garlic

R. Solv. (2) 1.7;1.9 0.6; 0.8 0.39;0.46 0.16; 0.17 20.7; 26.8 o10; 11.3 udl;4.0

Donoratico (0) – – – – – – –

D. Gibicar et al. / Environmental Research 109 (2009) 355–367 361

ARTICLE IN PRESS

Table 3 (continued )

Vegetable THg MeHg MeHg/THg Se

(mg/kg DW) (mg/kg FW) (mg/kg DW) (mg/kg FW) (mg/kg DW) (mg/kg FW)

Onion

R. Solv. (4) 3.774.8 0.570.7 0.4570.49 0.0670.05 36.8737.0 29.3724.7 3.873.0

Donoratico (0) – – – – – – –

Bean

R. Solv. (4) 0.870.5 0.270.1 0.1770.06 0.0470.02 34.6731.7 15.277.0 4.373.1

Donoratico (0) – – – – – – –

Radish

R. Solv. (1) 19.9 8.0 0.36 0.15 1.83 3.51 14.0

Donoratico (0) – – – – – – –

Celery

R. Solv. (2) 21.5; 26.9 3.3; 7.9 0.37; 0.67 0.06; 0.20 1.7; 2.5 o10; 24.1 udl; 7.1

Donoratico (0) – – – – – – –

Potato

R. Solv. (1) 1.5 0.3 0.16 0.03 10.6 20.5 4.2

Donoratico (0) – – – – – – –

Aubergine

R. Solv. (3) 0.770.5 0.170.1 0.2670.12 0.0470.03 52.5735.9 19.2711.9 2.371.2

Donoratico (0) – – – – – – –

Gourd

R. Solv. (1) 0.3 0.04 0.20 0.03 67.9 10.7 1.4

Donoratico (0) – – – – – – –

Plum

R. Solv. (1) 0.6 0.1 0.11 0.02 19.5 o10 udl

Donoratico (0) – – – – – – –

a Rosignano Solvay (number of samples).b Average7Std. dev.

D. Gibicar et al. / Environmental Research 109 (2009) 355–367362

edible marine species, which is 0.5 mg/kg FW. In the study ofStorelli et al. (2005) around 20% of analysed samples of stripedmullet caught from the Ionian and Adriatic Seas exceeded the EUlaw limit for mercury in seafood.

2.5. Exposure assessment

2.5.1. Exposure to mercury vapour in Rosignano Solvay

Exposure to elemental mercury by the general population andin occupational settings is primarily through inhaling mercuryvapours. Considering measured concentrations of TGM in air(Table 2), daily intake of inhaled atmospheric Hg in RosignanoSolvay was estimated at 0.06–2.0mg/day of which 0.05–1.6mg/dayis retained in the body. For the reference area, intake of Hg fromambient air has been estimated at 0.03–0.11mg/day (Table 5).Although the elemental Hg intake is estimated to be up to18 times higher in Rosignano Solvay than in the reference area,this intake is still relatively low and constitutes maximum up to50% of tolerable intake 4mg/day (WHO, 2003). Dental amalgamfor example is a much more significant source of Hg0 exposure.Estimates of daily intake from amalgam restorations range from 1to 27mg/day (ATSDR, 1999) with the majority of dental amalgamholders being exposed to less than 5mg mercury/day (WHO,2003). Of course there is a considerable variation in Hg intakefrom dental amalgam between individuals, primarily due to thenumber of amalgams present, gum chewing habits and bruxism.

By summing intake of elemental Hg from two sources: ambientHg coming from the MCCA plant (up to 50% PTDI) and Hg inhaledfrom evaporating amalgam fillings (up to 125% PTDI according toWHO (2003), the tolerable intake of Hg0 of 4mg/day can beexceeded (following worst case scenario for up to 175% of PTDI,Table 5).

2.5.2. Estimated dietary exposure to inorganic mercury

We estimated dietary exposure to I-Hg from consumptionof fish with concentrations of total mercury in the range of0.049–2.48 mg/kg of which up to 62.8% were in the form ofinorganic mercury. If we assume average consumption of 32 gof marine fish per day as reported by the EFSA’s journal (2004),the daily intake consists of up to 22.9mg (Table 5) or 16.4% of PTDIset by WHO (2003). Daily intake of inorganic mercury fromconsumption of locally grown vegetables was estimated at up to7.1mg for Rosignano Solvay and up to 5.0mg for the reference areaDonoratico, if consumption of 300 g of vegetable per day wasassumed (Table 5).

Considering intakes of inorganic Hg from other possiblesources, such as drinking water and food other than fish andvegetables taken from the literature (IPCS, 1991; WHO, 2003), thetotal PDI for I-Hg for an average person (70 kg body weight) inRosignano Solvay ranges from 0.052 to 0.481mg/kg body weight/day or 3–24% of tolerable intake of 2mg/kg body weight/day forI-Hg set by the WHO (2003). The intake of inorganic mercury

ARTICLE IN PRESS

Table 4Total and methylmercury (expressed as Hg) in muscles of marine organisms, determined by CV AAS, GC-ECD and CV AFS.

Fish species Weight (g) Sampling location THg (mg/kg) MeHg expressed

as Hg (mg/kg)

MeHg/THg (%)

Trachinus lineolatus Striped weever 140 1.5 km in front of Punta di Castinglioncello 0.185 0.185 100

Trigla lucerna Tub gurnard 120 1 mile km in front of Punta di Castinglioncello 0.218 0.122 56.0

Scorpaena Notata Scorpion fish 95 5.6 miles in front of the port Vada and Cecina 1.20 0.727 60.4

Scorpaena Notata Scorpion fish 105 5.6 miles in front of the port Vada and Cecina 0.260 0.129 49.4

Trigla lucerna Tub gurnard 75 5.6 miles in front of the port Vada and Cecina 0.049 0.022 45.5

Serranus scriba Painted comber 90 5.6 miles in front of the port Vada and Cecina 1.55 0.834 54.0

Uranoscopus scaber Stargazer 135 5.6 miles in front of the port Vada and Cecina 0.133 0.071 53.4

Conger conger Conger eel 100 5.6 miles in front of the port Vada and Cecina 0.424 0.202 47.6

Lichia amia Leerfish; jack 145 1.5 miles in front of the ‘‘fosso bianco’’ di Vada 0.170 0.107 63.0

Pomatomus saltatrix Blue fish 110 1.5 miles in front of the ‘‘fosso bianco’’ di Vada 0.454 0.290 63.9

Eriphia Verrucosa Crab 120 1.5 miles in front of the ‘‘fosso bianco’’ di Vada 0.508 0.292 57.4

Scorpaena Notata Scorpion fish 95 Castiglion della Pescaia 0.429 0.372 86.7

Scorpaena Notata Scorpion fish 105 Castiglion della Pescaia 0.502 0.415 82.7

Trachinus lineolatus Striped weever 340 Vicinity of Island Elba 0.855 0.605 70.7

Trigla lucerna Tub gurnard 400 Vicinity of Island Elba 0.129 0.093 72.1

Scorpaena Notata Scorpion fish 100 Vicinity of Island Elba 0.115 0.095 82.5

Uranoscopus scaber Stargaze 460 Vicinity of Island Elba 1.08 0.401 37.2

Serranus scriba Painted comber 90 Vicinity of Island Elba 0.680 0.680 100

Conger conger Conger eel 530 Vicinity of Island Elba 0.352 0.239 68.0

Seriola dumerili Amberjack 200 Vicinity of Island Elba 0.301 0.184 60.9

Trachinus lineolatus Striped weever 160 Market in Piombino 2.48 2.38 95.9

Trigla lucerna Tub gurnard 130 Market in Piombino 0.251 0.234 93.2

Trigla lucerna Tub gurnard 70 Market in Piombino 0.534 0.519 97.2

Serranus scriba Painted comber 95 Market in Piombino 0.860 0.845 98.3

Uranoscopus scaber Stargaze 135 Market in Piombino 1.98 1.90 96.0

Conger conger Conger eel 100 Market in Piombino 0.570 0.547 96.0

D. Gibicar et al. / Environmental Research 109 (2009) 355–367 363

Rosignano Solvay seems to be similar as in the reference areaDonoratico.

An important source of I-Hg that has not been included in thePDI estimation is consumption of mushrooms, which is veryfrequent in Italy. A variety of mushroom species have been shownto contain elevated levels of mercury (Bressa et al., 1988; Kalacet al., 1991). The extent of bioaccumulation of mercury seems tobe species-dependent (Kalac et al., 1991); the edible mushroomPleurotus ostreatus has been found to bioaccumulate up to140 times the concentration in the soil (Bressa et al., 1988).

There are also a number of other possible pathways for non-occupational exposure to inorganic forms of mercury. Theseinclude playing on or in contaminated surface soils; playing withliquid mercury from broken electrical switches, thermometers,barometers, blood pressure monitors, etc. Exposure from drinkingwater is usually minor (WHO, 2003).

2.5.3. Estimated exposure to methylmercury

In 2003 the Joint WHO/FAO Expert Committee on FoodAdditives (JECFA) recommended a provisional tolerable weeklyintake (PTWI) for methylmercury of 1.6mg/kg body weight/week(equivalent to 0.23mg methylmercury/kg body weight/day (PTDI))in order to sufficiently protect the developing foetus (WHO, 2003).The RfD for MeHg set by the US EPA (2001) is even lower,0.1mg methylmercury/kg body weight day.

In Rosignano Solvay, the intake of methylmercury has beenestimated at 0.70–76.26mg/day from consumption of local marinefish and only up to 0.10mg/day from consumption of locally grownvegetables (Table 5). An average adult person in Rosignano Solvay(70 kg) consuming on average 32 g fish from the local marketconsumes and retains up to 1.089mg MeHg/kg body weight/day.The calculated PDI/PTDI ratio for MeHg ranges up to 474% andexceeds the toxicologically tolerable value set by JEFCA (2003).The elevated safety limit from consumption of marine fish fromAdriatic and Ionian Seas was observed also by Storelli et al. (2005).

Fish are an important source of proteins, vitamin D, selenium,long-chain n-3 fatty acids eicosapentaenoic acid (EPA), docosa-pentaenoic acid (DPA) and docosahexaenoic acid (DHA). Sincesome fish contain elevated MeHg concentrations and otherharmful substances, such as polychlorinated biphenyls anddioxins, many countries already advise that susceptible popula-tion strata should avoid consummation of certain type of fish.Methylmercury in fish is primarily bound to the amino acids infish muscle; thus, skinning, trimming and cooking do notsignificantly reduce MeHg concentrations (Kris-Etherton et al.,2002).

2.5.4. Total mercury exposure in Rosignano Solvay

The PTDI for total mercury has been estimated by JEFCA (2003)at 0.71mg/kg body weight/day (equivalent to 5mg/kg body weight/week). In Rosignano Solvay the PDI for total mercury for anaverage 70-kg adult has been estimated at 0.075–1.389mg/kg body weight/day, which is similar to the intake in the referencearea (0.075–1.332mg/kg body weight/day). According to our esti-mations (Table 5), the PDI/PTDI ratio for total mercury inhaled andingested from various sources (air, amalgam, fish, vegetables,drinking water and food other than vegetables and fish) rangedfrom 11% to 196% in Rosignano Solvay and 11–188% in thereference area.

Estimated PDI/PTDI ratios for total mercury in RosignanoSolvay were: up to 4% from inhalation of ambient air, up to 14%from ingesting local vegetables and up to 160% from consuminglocal marine fish. Estimated PDI/PTDI ratios for total mercury inDonoratico as the reference area were: up to 0.3% from inhalationof ambient air, up to 10% from ingesting local vegetables and up to160% from consuming local marine fish.

According to EFSA’s report from 2004, the average estimateddaily intake of total mercury in Italy is 8.6mg/day or 0.123mg/kgbody weight/day for an average 70-kg adult, which corresponds to17.2% of the PTDI. EFSA’s estimation is based on lower Hg

ARTICLE IN PRESS

Table 5Estimated range of daily intakes of inorganic mercury, elemental mercury, methylmercury and total mercury in mg for inhabitants of Rosignano Solvay and the reference

area.

Medium Probable daily intake (PDI) in mg/day

Hg vapour Inorganic Hg MeHg Total Hg

Aira

Rosignano Solvay 0.06–2.0a 0b 0b 0.06–2.00

Reference area 0.03–0.11 0 0 0.03–0.11

Local vegetablesc

Rosignano Solvay 0 0.00–7.10 0.00–0.10 0.00–7.20

Reference area 0 0.01–5.04 0.00–0.09 0.01–5.13

Fishd 0 0.00–22.9 0.70–76.16 1.57–79.36

Amalgame 0.0–5.0 0 0 0.0–5.0

Drinking waterf 0 0.05 0 0.05

Other foodg 0 3.6 0 3.6

Total

Rosignano Solvay 0.06–7.00 3.65–33.66 0.70–76.26 5.28–97.2

Reference area 0.03–5.11 3.66–31.61 0.70–76.26 5.26–93.3

Total PDI in mg/kg bw/dayh

Rosignano Solvay 0.001–0.100 0.052–0.481 0.010–1.089 0.075–1.39

Reference area 0.000–0.073 0.052–0.452 0.010–1.089 0.075–1.33

Provisional tolerable daily intake (PTDI)

4mg/dayi 2mg/kg bw/dayi 0.23mg/kg bw/dayj 0.71mg/kg bw/dayj

PDI/PTDI (%)

Rosignano Solvay (%) 2–175 3–24 4–474 11–196

Reference area (%) 1–128 3–23 4–474 11–188

Daily intakes from sources of Hg, such as dental amalgam, drinking water and food other than fish and vegetables were taken from the literature.a Assumes a daily respiratory volume of 20 m3, and air concentration of 2.8–100 ng/m3 for Rosignano Solvay and 1.5–5.5 ng/m3 for Donoratico.b For the purposes of comparison, it is assumed that the atmospheric concentrations of species of mercury other than mercury vapour are negligible.c Assumes a daily intake of 300 g of vegetables with concentration of total Hg 0.01–24 ng/g FW and MeHg 0.002–0.348 ng/g FW for Rosignano Solvay and concentration

of total Hg 0.03–17.1 ng/g FW and MeHg 0.001–0.29 ng/g FW for Donoratico. Concentration of inorganic Hg is calculated as (conc. of total Hg�conc. of MeHg).d Assumes a daily intake of 32 g of fish with concentration of 0.022–2.38mg MeHg/g FW and 0.049–2.48mg total Hg/g FW. Concentration of inorganic Hg is calculated as

(conc. of total Hg�conc. of MeHg) and ranges from 0.00–0.716mg I-Hg/g FW.e Assumes a daily intake of 1.2–5mg from dental amalgam, reported by WHO (2003). Daily intake of up to 27mg from dental amalgam has been reported by ATSDR

(1999).f Assumes a daily intake of 0.05mg of inorganic Hg (IPCS, 1991; WHO, 2003).g Assumes a daily intake of 3.6mg of inorganic Hg (IPCS, 1991; WHO, 2003).h Assumes an average body weight of 70 kg.i PTDI according to WHO (2003).j PTDI according to JEFCA (2003).

D. Gibicar et al. / Environmental Research 109 (2009) 355–367364

concentrations in fish and molluscs (0.07–0.45mg/g) which weregathered from available data from 1997. The average intake oftotal mercury by an adult in the Member States of EFSA’sreport is estimated at 5.53mg/day or 0.079mg/kg body weight/day for an average 70-kg adult. The EFSA’s data show that for afood item like fish the variation of mean consumption in differentcountries across Europe is very high and the variation in foodconsumption could result in exposures that vary by a factorof 10.

3. Discussion

MCCA plants produce chlorine, thus the emission plume maycontain a mixture of Hg0, Cl2 and HgCl2. The main sources ofgaseous Hg in the caustic-chlorine industry are ventilation airoutlets from cell rooms, H2 gas outlets, Hg evasion from wastewater and solid wastes. Emission of mercury from the MCCA plant

in Rosignano Solvay was measured by differential absorptionLIDAR (DIAL) technique in 2003 by Gronlund et al. (2005). Theyestimated the emission of Hg from the plant to be around20–54 g/h, which was similar to the earlier measurements madein 1990 (around 43 g/h) by Ferrara et al. (1992). The measured fluxdepends on ambient temperature and wind speed, since theseparameters affect the temperature of the cells and the re-emissionprocesses from contaminated soil around the cell and from spillsdistributed in the cell house structures (Gronlund et al., 2005;Bennett et al., 2006).

Concentrations of ambient air measured in Rosignano Solvay in2002 were relatively low: 8.0–8.7 ng/m3 in summer and2.8–4.2 ng/m3 in winter, although at particular meteorologicalconditions concentrations of up to 100 ng/m3 were observed.Background mercury levels (2 ng/m3) were reached already 3 kmfrom the MCCA plant. In Poland for example, 1800 m from the cellof the MCCA plant in Tarnow, mean measured TGM concentrationin ambient air were 16 ng/m3 (Jarosinska et al., 2006).

ARTICLE IN PRESS

D. Gibicar et al. / Environmental Research 109 (2009) 355–367 365

We found most of the mercury in the atmosphere around theMCCA plant in Rosignano Solvay in its elemental form, whichcirculates in the atmosphere and can be widely dispersed andtransported over thousands of kilometres affecting soils andaquatic environments in remote areas. We observed only around4.2% of total emitted Hg present in the form of HgII, which is incontrast to Hg0 easily deposited to ground and vegetation byprecipitation or dry deposition. Only around 1.0% of total Hg waspresent in the form of TPM. The deposition of emitted Hg dependson the amount of HgII and TPM emitted, rather than the totalmercury emitted. The results of modelling of emitted Hg showedthat only 14% (or 6.8 kg/yr) of total emitted Hg were depositedwithin 5 km from the plant, while the remaining 86% weredispersed and transported away from the area, showing itscharacteristics of a global pollutant.

As mercury cycles between atmosphere, pedosphere andhydrosphere, undergoes many complex chemical and physicaltransformations such as biotic and abiotic methylation orHg0 oxidation which may increase toxicity and bioavailability of themetal (Lindqvist et al., 1991; Stein et al., 1996). Mercury depositedto soil is affected by various chemical and biological transforma-tion processes such as Hg0 oxidation, HgII reduction or methyla-tion, depending on several soil parameters (pH, temperature, soilhumic substances content, the nature of organic matter and itsquantity, etc.) (Weber, 1993; Remy et al., 2006; Wang et al., 2003).Methylation and demethylation processes involve activity of SO4-reducing bacteria, methanogens, photoreduction (Costa andLiss, 2000) and other abiotic processes in soil. If Hg0 is drydeposited to soil, it has the potential to be converted to reactiveHg and sequestered in the soil. If HgII is dry deposited tosoils, it has the potential to be reduced and re-emitted throughvolatilization. The chemistry of either oxidation or reduction mayprevail. In case of prevailing oxidation, a larger metal retentionmay occur after some time, since it is the reduced speciesthat returns to the atmosphere through volatilization and notthe oxidized one. Since atmospheric levels of mercury measuredin Rosignano Solvay were only two–three times higher thanthose measured at the reference station, as expected alsosurface soil from the Rosignano Solvay area contained relativelylow levels of total mercury (30.1–2919mg/kg DW) and methyl-mercury (0.02–3.88mg/kg DW). Similar concentrations of THg insoil from the same area were found also by Maserti and Ferrara in1991.

Assimilation through plants plays a major role in the entry ofmercury into terrestrial food chains (Barghigiani and Ristori, 1994;WHO, 1989). The assimilation of mercury by plants does notdepend only on its concentration in soil, but also on the ratio ofsoil and air mercury contamination, on bio-chemical conditions ofsoil and on meteorological conditions. Vascular plants accumulateHg (a) through the roots from soil as ionic Hg or dissolved gaseousHg (DGM); (b) through the stomata from the atmosphere,particularly as Hg0; (c) through foliar adsorption of divalent,reactive gaseous mercury and particulate Hg (Shaw and Panigrahi,1986; Lindberg et al., 1992; Bishop et al., 1998). Plants also emitmercury vapour from leaves (Ericksen et al., 2003). Vegetation canthus act as a source or a sink of atmospheric Hg. Data frommonitoring forested ecosystems indicates that dry deposition toforest canopies is a significant removal process for both atmo-spheric Hg0 and MeHg species (Lindberg et al., 1991; Iverfeldt, 1991;Hultberg et al., 1994; Munthe et al., 1995). The assimilation from soilby roots depends on soil type, content of humic acid, microbiologicalactivity, pH and redox potential. The assimilation through leavesdepends on type of plant, air contamination and atmospheric aerosoldeposition. Bryophytes and lichens have no roots and assimilatemercury only from air and water and can be used as activebiomonitors (Fernandez et al., 2002; Bargagli and Barghigiani, 1991;

Bargagli et al., 1987; Maserti and Ferrara, 1991). Mercury assimilationby some vascular plants is considered to be negligible, whereasothers bioconcentrate mercury (e.g., Pinus sp., About et al., 2001).These plants have been used under experimental conditions fordecontamination of polluted soils. It is known that certain species,such as carrots, lettuce and mushroom in particular, are likely toassimilate more mercury than others growing at the same location.In our study the highest concentrations of mercury (expressed on adry weigh) were observed in basil, beet, celery, parsley, and mint.Vegetables grown in the area around the MCCA plant in RosignanoSolvay contained slightly higher levels of total and methylmercurycompared to the vegetables grown 20 km south, but the differencewas not statistically significant and the number of samples wasrelatively low.

Another contribution to the mercury pollution by the MCCAplant in Rosignano Solvay is the mercury discharged through aneffluent channel into adjacent coastal seawater. Ferrara et al.(2001) estimated the release of mercury in the coastal seawaterthrough a discharge ditch of the MCCA plant in Rosignano Solvayto be about 400–500 kg/yr at the present chlorine production.They also estimated that from the 1950s to 1973 12–14 ton/yr ofmercury flowed into the sea through a discharge ditch, whichwere incorporated into the surrounding sediments. Once in theaquatic ecosystem, part of the inorganic mercury can be micro-bially converted into methylmercury and taken up by aquaticorganisms. It is known that fish accumulate mercury directly fromfood and from the surrounding water (Rainbow, 1985) and canbioconcentrate large amounts of this metal. Methylmercuryconcentration in fish depends on the feeding habits of the fish,mercury concentrations in tissues of its prey, the fish’s age andposition in the food chain. The highest concentrations are found inlarge and old predatory fish, such as sharks, swordfish, tuna andpike. Mediterranean Sea constitutes beside the anthropogenicinput the richest natural reserve of Hg. Deep environment of theMediterranean Sea, the high temperature and the absence of solarradiation favour a high methylation rate (Bacci, 1989) resulting inbioconcentration in predator species and biomagnificationthrough the benthic food chains. Many studies have describedelevated levels of mercury in marine organisms from theMediterranean Sea (Renzoni and Baldi, 1975; Hornung et al.,1993; Storelli et al., 2002, 2005, 2007).

The elevated Hg content in marine fish observed during ourstudy cannot be directly associated with the emissions of Hg fromthe MCCA plant in Rosignano Solvay, since many other (naturaland anthropogenic) sources of Hg in the Mediterranean exist.However, for freshwater fish, an assumption is made that anincrease in local deposition leads to a linear increase in mercurylevels in local freshwater fish. Fish and sediments collected inrivers or canals downstream the MCCA plants contain higherHg levels compared to the samples collected upstream (Scerboet al., 2005; Raldua et al., 2007). Moreover, histopathologicalanalysis showed that fish downstream the MCCA plants havesignificantly higher prevalence of liver pathologies (Raldua et al.,2007).

We performed simple exposure and risk assessment forelemental, inorganic, methyl and total mercury, drawn onmeasurements conducted for this study, as well as on measure-ments and intake estimates drawn from the literature. Weconcluded that following worst case scenario, people living nearthe MCCA plant in Rosignano Solvay inhale up to 2.0mg Hg0/dayfrom the ambient air (or 50% of the PTDI according to WHO(2003), while people living in the reference area Donoratico inhaleup to 0.11mg Hg0/day. Dental amalgam is a much more significantsource of exposure to elemental mercury, whose contribution canexceed the provisional tolerable intake of 4mg/day. Estimates ofdaily intake from amalgam restorations range from 1 to 27mg/day

ARTICLE IN PRESS

D. Gibicar et al. / Environmental Research 109 (2009) 355–367366

(ATSDR, 1999) with the majority of dental amalgam holders beingexposed to less than 5mg mercury/day (WHO, 2003). Of coursethere is a considerable variation in Hg intake from dentalamalgam between individuals, primarily due to the number ofamalgams present, gum chewing habits and bruxism. By sum-ming intake of elemental Hg from two sources: ambient Hgcoming from the MCCA plant (up to 50% PTDI) and Hg inhaledfrom evaporating amalgam fillings (up to 125% PTDI according toWHO (2003), the tolerable intake of Hg0 of 4mg/day can beexceeded (following worst case scenario for up to 175% of PTDI,Table 5).

Estimated exposure to inorganic mercury by ingestion of localvegetables, fish and other food in inhabitants of Rosignano Solvaywas similar as in the reference population (up to 33.66mg/day or24% of PTDI) and showed no unacceptable risks to human health.In contrary, estimated exposure to methylmercury seems to posean elevated risk of possible adverse effects of methylmercuryexposure, since marine fish from this area contain elevated levelsof methylmercury. Considering average consumption of 32 g offish per day, the PDI/PTDI ratio can be easily exceeded (up to474%). Our estimated daily intakes of mercury relay on averagedaily intake values and as such underestimate intakes in the caseof higher consumption rates. Diets vary considerably amongindividuals in the study population, so the variation in foodconsumption could result in exposures that vary by a factor of 10.

Human health risk assessment for mercury emitted from theMCCA plants has been performed also by the organisationrepresenting chlorine industry named Euro Chlor (2003), basedupon the data from the immediate vicinity of mercury-basedplants operating according to current Best Available Technologyunder the terms of the Integrated Pollution Prevention andControl (IPPC) Directive 96/61/EC. Their results of worst casescenario for human health risk assessment for inorganic andelemental mercury, calculated from the DI/TDI ratio (daily Hgintake (DI) ¼ 3.12mg/kg body weight/day; tolerable daily intake(TDI) ¼ 4mg/kg body weight/day), showed no unacceptable risksto human health. For the aquatic compartment and the sedimentsthe predicted exposure concentrations/predicted no-effect con-centrations (PEC/PNEC) ratios were less than one, showing nounacceptable risk. The PEC/PNEC ratios for the terrestrialcompartment varied from less then one for soil organisms tomore than one for terrestrial predators. For inorganic and organicmercury worst case scenario of risk assessment for the marineenvironment in the North Sea (OSPAR) area, including variousecotoxicology studies, showed DI/TDI ratio higher than one for seabirds as aquatic predators (Euro Chlor, 1999). It was concludedthat mercury is a persistent, toxic and bioaccumulable chemical.It’s potential for secondary poisoning (food-chain biomagnifica-tion) and for long-range transport, makes it a high prioritychemical for emission control. However, due to its naturaloccurrence, mercury will remain ubiquitous in the environment.

4. Conclusions

The aim of our study was to provide analysis of potentialmercury risks in an area near a mercury cell chlor-alkali plant inRosignano Solvay in Italy. The main conclusion is that the impactof the mercury cell chlor-alkali plant in Rosignano Solvay on thelocal terrestrial environment is restricted only to the closesurrounding area. Only 14% of emitted gaseous Hg from themercury cell chlor-alkali plant is deposited within 5 km from thesource. The remaining 86% are dispersed and transported awayfrom that area increasing the atmospheric pool of mercury.Exposure assessment performed by calculating ratios of probabledaily intake and provisional tolerable daily intake of mercury for

various exposure pathways showed no unacceptable risks tohuman health for elemental and inorganic mercury, except forindividuals with higher number of amalgam fillings. In contrary,the PDI/PTDI ratio for methylmercury and total mercury exceededthe toxicologically tolerable value due to the potential consump-tion of local marine fish. Exceeded PDI/PTDI ratio for methylmer-cury and total mercury from consumption of contaminated fish isnot directly associated with the emissions of mercury from theMCCA plant in Rosignano Solvay, but is a product of anthropogenicand natural sources of Hg present in the Mediterranean. Weconclude that any exceeded tolerable daily intake values are dueto other sources of mercury, rather than the mercury cell chlor-alkali plant.

Acknowledgments

This work was supported by the 5FP EU project EMECAP(European Mercury Emission from Chlor-Alkali Plants, QLK4-CT-2000-00489), the Ministry of Education, Science and Sport of theR Slovenia (Programme P1-0143) and by financial support by theyoung researchers programme. Part of the programme was alsoco-financed by the Centre of Excellence for EnvironmentalTechnologies funded by the EU structural funds and the IAEACRP 13250/RD. The authors wish to thank Ana Nusa Prosenc forhelp with Hg and Se analysis by neutron activation analysis.

References

About, J.R., Fernandez, J.A., Carballeira, A., 2001. Sampling optimization, at sitescale, in contamination monitoring with moss, pine and oak. Environ. Pollut.115, 313–316.

ATSDR, 1999. Toxicological Profile for Mercury. Agency for Toxic Substances andDisease Registry, US Department of Health and Human Services, Public HealthService, Atlanta, GA.

Bacci, E., 1989. Mercury in the Mediterranean. Mar. Pollut. Bull. 20, 59–63.Baldi, F., Bargagli, R., 1984. Mercury pollution in marine sediment near a chlor-

alkali plant: distribution and availability of the metal. Sci. Total Environ. 39,15–26.

Baldi, F., D’Amato, M.L., 1986. Mercury pollution in marine sediment cores nearcinnabar deposits and a chlor-alkali plant. Sci. Total Environ. 57, 111–120.

Bargagli, R., Iosco, F.P., Barghigiani, C., 1987. Assessment of mercury dispersal in anabandoned mining area by soil and lichens analysis. Water Air Soil Pollut. 36,219–225.

Bargagli, R., Barghigiani, C., 1991. Lichen biomonitoring of mercury emission anddeposition in mining, geothermal and volcanic areas of Italy. Environ. Monit.Assess. 16, 265–275.

Barghigiani, C., Bauleo, R., 1992. Mining area environmental mercury assessmentusing Abies alba. Bull. Environ. Contam. Toxicol. 49, 31–36.

Barghigiani, C., Ristori, T., Biagi, F., De Ranieri, S., 2000. Size related mercuryaccumulations in edible marine species from an area of the northernTyrrhenian Sea. Water Air Soil Pollut. 124, 169–176.

Barghigiani, C., De Ranieri, S., 1992. Mercury content in different size classes ofimportant edible species of the northern Tyrrhenian Sea. Mar. Pollut. Bull. 24,114–116.

Barghigiani, C., Pellegrini, D., D’Ulivo, A., De Ranieri, S., 1991. Mercury assessmentand its relation to selenium levels in edible species of the northern TyrrhenianSea. Mar. Pollut. Bull. 22, 406–409.

Barghigiani, C., Ristori, T., 1994. Mercury levels in land products of Mt. Amiata(Tuscany, Italy). Arch. Environ. Contam. Toxicol. 26, 329–334.

Bennett, M., Edner, H., Gronlund, R., Sjoholmb, M., Svanberg, S., Ferrara, R., 2006.Joint application of Doppler Lidar and differential absorption lidar to estimatethe atomic mercury flux from a chlor-alkali plant. Atmos. Environ. 40,664–673.

Bishop, K.H., Lee, Y.-H., Munthe, J., Dambrine, E., 1998. Xylem sap as a pathway fortotal mercury and methylmercury transport from soils to tree canopy in theboreal forest. Biogeochemistry 40, 101–113.

Bressa, G., Cima, L., Costa, P., 1988. Bioaccumulation of mercury in the mushroomPleurotus ostreatus. Ecotoxicol. Environ. Saf. 16, 85–89.

Byrne, A.R., Kosta, L., 1974. Simultaneous neutron-activation determination ofselenium and mercury in biological samples by volatilization. Talanta 21,1083–1090.

Byrne, A.R., Skreblin, M., Falnoga, I., Al-Sabti, K., Stegnar, P., Horvat, M., 1995.Mercury and selenium: perspectives from Idrija. Acta Chim. Slov. 42, 175–198.

Clarkson, T.W., 1997. The toxicology of mercury. Crit. Rev. Clin. Lab. Sci. 34,369–440.

ARTICLE IN PRESS

D. Gibicar et al. / Environmental Research 109 (2009) 355–367 367

Costa, M., Liss, P., 2000. Photoreduction and evolution of mercury from seawater.Sci. Total Environ. 261, 125–135.

Denby, B., Pacyna, J., 2004. Deliverable D5.4. EMECAP. Final report to the EuropeanCommission, Brussels, Belgium, 2004.

DEFRA, 2002. Soil Guideline Values for Inorganic Mercury Contamination.Department for Environment, Food and Rural Affairs.

EFSA (European Food Safety Authority), 2004. SCOOP 3.2.11—intake of As, Cd, Pband Hg. EFSA J. 34, 1–14 Available at /http://www.mhlw.go.jp/shingi/2004/08/dl/s0817-2k2.pdfS.

Ericksen, J.A., Gustin, M.S., Schorran, D.E., Johnson, D.W., Lindberg, S.E., Coleman,J.S., 2003. Accumulation of atmospheric mercury in forest foliage. Atmos.Environ. 37, 1613–1622.

Euro chlor mercury risk assessment, Local scenario, an overview, 2003. Availableat: /http://www.chem.unep.ch/Mercury/2003-ngo-sub/Eurochlor/Euro%20Chlor%20Hg%20RA%20Local%20Scenario%20Summary.docS.

Euro Chlor, 1999. Euro chlor risk assessment for the marine environmentOSPARCOM Region—North Sea, Mercury. Available at: /http://www.eurochlor.org/HgS.

Euro Chlor, 2005. Chlorine Industry Review 2004–2005. Euro Chlor, Brussels.Available at: /http://www.eurochlor.orgS.

Falnoga, I., Tusek-Znidaric, M., Horvat, M., Stegnar, P., 2000. Mercury, selenium, andcadmiun in human autopsy samples from Idrija residents and mercury mineworkers. Environ. Res. Sect. A 84, 211–218.

Fernandez, J.A., Ederra, A., Nunez, E., Martynez-Abaigar, J., Infante, M., Heras, P., etal., 2002. Biomonitoring of metal deposition in northern Spain by mossanalysis. Sci. Total Environ. 300, 115–127.

Ferrara, R., Maserti, B.E., Paterno, P., 1989. Mercury distribution in marine sedimentand its correlation with the Posidonia oceanica prairie in a coastal area affectedby a chlor-alkali complex. Toxicol. Environ. Chem. 22, 131–134.

Ferrara, R., Maserti, B.E., Edner, H., Ragnarsson, P., Svanberg, S., Wallinder, E., 1992.Mercury emissions into the atmosphere from a chlor-alkali complex measuredwith the lidar technique. Atmos. Environ. 26A (7), 1253–1258.

Ferrara, R., Lanzillotta, E., Ceccarini, C., 2001. Dissolved gaseous mercuryconcentration and mercury evasional flux from seawater in front of a chlor-alkali plant. Environ. Technol. 22 (8), 971–978.

Gronlund, R., Sjoholm, M., Weibring, P., Edner, H., Svanberg, S., 2005. Elementalmercury emissions from chlor-alkali plants measured by lidar techniques.Atmos. Environ. 39, 7474–7480.

Hornung, H., Krom, M.D., Cohen, Y., Bernhard, N., 1993. Trace metal content in deepwater sharks from the eastern Mediterranean Sea. Mar. Biol. 115, 331–338.

Horvat, M., Byrne, A.R., May, K., 1990. A modified method for the determination ofmethylmercury by gas chromatography. Talanta 37 (2), 207–212.

Horvat, M., Liang, L., Bloom, N.S., 1993. Comparison of distillation with othercurrent isolation methods for the determination of methylmercury compoundsin low level environmental samples, Part I. Sediment. Anal. Chim. Acta 281,135–152.

Horvat, M., Lupsina, V., Pihlar, B., 1991. Determination of total mercury in coal flyash by gold amalgamation cols vapour atomic absorption spectrometry. Anal.Chim. Acta 243, 71–79.

Horvat, M., May, K., Stoeppler, M., Byrne, A.R., 1988. Comparative studies ofmethylmercury determination in biological and environmental samples. Appl.Organomet. Chem. 2, 515–524.

Horvat, M., Mandic, Liang, L., Bloom, N.S., Padberg, S., Lee, Y.H., Hintelmann, H.,Benoit, J., 1994. Certification of methylmercury compounds concentration inmarine sediment reference material, IAEA-356. Appl. Organomet. Chem. 8,533–540.

Hultberg, H., Iverfeldt, A., Lee, Y.-H., 1994. Methylmercury input–output andaccumulation in forested catchments and critical loads for lakes in south-western Sweden. In: Watras, C.J., Huckabee, J.W. (Eds.), Mercury Pollution:Integration and Synthesis. Lewis Publishers, Chelsea, MI, pp. 313–322.

IPCS, 1991. Inorganic Mercury. World Health Organization, International Pro-gramme on Chemical Safety, Geneva (Environmental Health Criteria 118).

Iverfeldt, A., 1991. Mercury in forest canopy through fall water and its relation toatmospheric deposition. Water Air Soil Pollut. 56, 553–564.

Jarosinska, D., Barregard, L., Biesiada, M., Muszynska-Graca, M., Dabkowska, B.,Denby, B., Pacyna, J., Fudala, J., Zielonka, U., 2006. Urinary mercury in adults inPoland living near a chlor-alkali plant. Sci. Total Environ. 368, 335–343.

Joint FAO/WHO Expert Committee on Food Additives (JEFCA), World HealthOrganization (WHO), 2003. Sixty-first meeting 10–19 June 2003, summary andconclusions. Available on-line: /ftp://ftp.fao.org/es/esn/jecfa/jecfa61sc.pdfS.

Kalac, P., Burda, J., Staskova, I., 1991. Concentrations of lead, cadmium, mercury,and copper in mushrooms in the vicinity of a lead smelter. Sci. Total Environ.105, 109–119.

Kim, K.H., Lindberg, S.E., 1995. Design and initial tests of a dynamic enclosurechamber for measurments of vapour-phase mercury fluxes over soils. Sci. TotalEnviron. 80, 1059–1068.

Kosta, L., Byrne, A.R., Zelenko, V., 1975. Correlation between selenium and mercuryin man following exposure to inorganic mercury. Nature 254, 238–239.

Kris-Etherton, P.M., Harris, W.S., Appel, L.J., for the Nutrition Committee, 2002. Fishconsumption, fish oil, omega-3 fatty acids, and cardiovascular disease.Circulation 106, 2747–2757.

Liang, L., Horvat, M., Bloom, N.S., 1994a. An improved speciation method formercury by GC/CVAFS after aqueous phase ethylation and room temperatureprecollection. Talanta 41 (3), 371–379.

Liang, L., Bloom, N.S., Horvat, M., 1994b. Simultaneous determination of mercuryspeciation in biological materials by GC/CVAFS after ethylation and room-temperature precollection. Clin. Chem. 40, 602–607.

Lindberg, S.E., Turner, R.R., Meyers, T.P., Taylor Jr., G.E., Schroeder, W.H., 1991.Atmospheric concentrations and deposition of airborne Hg to Walker branchwatershed. Water Air Soil Pollut. 56, 577–594.

Lindberg, S.E., Meyers, T.P., Taylor Jr., G.E., Turner, R.R., Schroeder, W.H., 1992.Atmosphere–surface exchange of mercury in a forest: results of modeling andgradient approaches. J. Geophys. Res. 97 (D2), 2519–2528.

Lindqvist, O., Johansson, K., Aastrup, M., Andersson, A., Brinkmark, L., Hovsenius,G., Hakanson, L., Iverfeldt, A., Meili, M., Timm, B., 1991. Mercury in the Swedishenvironment: recent research on causes, consequences and correctivemethods. Water Air Soil Pollut. 55, 1–261.

Logar, M., Horvat, M., Falnoga, I., Stibilj, V., 1999. A methodological study ofmercury speciation using Dogfish liver CRM (DOLT-2). Fres. J. Anal. Chem. 366,452–460.

Maserti, B.E., Ferrara, R., 1991. Mercury in plants, soil and atmosphere near a chlor-alkali complex. Water Air Soil Pollut. 56, 15–20.

McLaughlin, M.J., Parker, D.R., Clarke, J.M., 1999. Metals and micronutrients—foodsafety issues. Field Crops Res. 60, 143–163.

Munthe, J., Hultberg, H., Iverfeldt, A., 1995. Mechanisms of deposition ofmethylmercury and mercury to coniferous forests. Water Air Soil Pollut. 80,363–371.

Pacyna, E.G., Pacyna, J.M., Fudala, J., Strzelecka-Jastrzab, E., Hlawiczka, S., Panasiuk,D., 2006. Mercury emissions to the atmosphere from anthropogenic sources inEurope in 2000 and their scenarios until 2020. Sci. Total Environ. 370, 147–156.

Rainbow, P.S., 1985. The biology of heavy-metals in the sea. Int. J. Environ. Stud. 25,195–211.

Raldua, D., Dıez, S., Bayona, J.M., Barcelo, D., 2007. Mercury levels and liverpathology in feral fish living in the vicinity of a mercury cell chlor-alkalifactory. Chemosphere 66 (7), 1217–1225.

Reilly, C., 1996. Selenium in Food and Health. Blackie Academic & Professional,London, Weinheim, New York, 323pp.

Remy, S., Prudent, P., Probst, J.L., 2006. Mercury speciation in soils of theindustrialised Thur River catchment (Alsace, France). Appl. Geochem. 21 (11),1855–1867.

Renzoni, A., Bacci, E., Falciai, L., 1973. Mercury concentration in the water,sediments and fauna of an area of the Tyrrhenian Coast. Rev. Intern. Oceanogr.Med. XXXI–XXXII, 17–45.

Renzoni, A., Baldi, F., 1975. Osservazioni sulla distribuzione di mercurio nella faunadel Mar Ligure e del Mar Tirreno. Acqua e Aria 8, 597–602.

Rossi, A., Pellegrini, D., Belcari, P., Barghigiani, C., 1993. Mercury in Eledone cirrhosafrom the northern Tyrrhenian Sea: contents and relations with life cycle. Mar.Pollut. Bull. 26, 683–686.

Scerbo, R., Ristoria, T., Stefaninia, B., De Ranieri, S., Barghigiani, C., 2005. Mercuryassessment and evaluation of its impact on fish in the Cecina river basin(Tuscany, Italy). Environ. Pollut. 135, 179–186.

Shaw, B.P., Panigrahi, A.K., 1986. Uptake and tissue distribution of mercury in someplant species collected from a contaminated area in India: its ecologicalimplications. Arch. Environ. Contam. Toxicol. 15, 439–446.

Stein, E.D., Cohen, Y., Winer, A.M., 1996. Environmental distribution and trans-formation of mercury compounds. Crit. Rev. Environ. Sci. Technol. 26, 1–43.

Storelli, M.M., Giacominelli-Stuffler, R., Marcotrigiano, G.O., 2002. Total andmethylmercury residues in cartilaginous fish from Mediterranean Sea. Mar.Pollut. Bull. 44, 1354–1358.

Storelli, M.M., Storelli, A., Giacominelli-Stuffler, R., Marcotrigiano, G.O., 2005.Mercury speciation in the muscle of two commercially important fish, hake(Merluccius merluccius) and striped mullet (Mullus barbatus) from theMediterranean Sea: estimated weekly intake. Food Chem. 89, 295–300.

Storelli, M.M., Barone, G., Garofalo, R., Marcotrigiano, G.O., 2007. Metals andorganochlorine compounds in eel (Anguilla anguilla) from the Lesina lagoon,Adriatic Sea (Italy). Food Chem. 100, 1337–1341.

US EPA (Environmental Protection Agency), 2001. Available at URL: /http://www.epa.gov/mercury/exposure.htmS.

Wang, D., Shi, X., Wei, S., 2003. Accumulation and transformation of atmosphericmercury in soil. Sci. Total Environ. 304, 209–214.

Wangberg, I., Edner, H., Ferrara, R., Lanzillotta, E., Munthe, J., Sommar, J., Sjoholm,M., Svanberg, S., Weibring, P., 2003. Atmospheric mercury near a chlor-alkaliplant in Sweden. Sci. Total Environ. 304, 29–41.

Weber, J.H., 1993. Review of possible paths for abiotic methylation of mercury (II)in the aquatic environment. Chemospere 26, 2063–2072.

WHO, 1989. Environmental Health Criteria 86: Mercury-Environmental Aspects.World Health Organisation, Geneva.

WHO, 2003. Concise International Chemical Assessment Document 50. ElementalMercury and Inorganic Mercury Compounds: Human Health Aspects. IPCS,WHO, Geneva, Switzerland.

Xiao, Z.F., Munthe, J., Schroeder, W.H., Lindqvist, O., 1991. Vertical fluxes of volatilemercury over forest soil and lake surfaces in Sweden. Tellus 43B, 267–279.