PCBs in central vietnam coastal lagoons levels and trends in dynamic environments

PCBs in Central Vietnam coastal lagoons: Levels and trendsin dynamic environments Silvia Giuliania,⇑, Rossano Piazzab,c, Luca Giorgio Belluccia, Nguyen Huu Cud, Marco Vecchiatob,c, Stefa

PCBs in Central Vietnam coastal lagoons: Levels and trends

in dynamic environments

Silvia Giuliania,⇑, Rossano Piazzab,c, Luca Giorgio Belluccia, Nguyen Huu Cud, Marco Vecchiatob,c,

Stefania Romanoa, Cristian Mugnaia, Dang Hoai Nhond, Mauro Frignania

a CNR-Istituto di Scienze Marine, Via Gobetti 101, 40129 Bologna, Italy

b Dipartimento di Scienze Ambientali, Università di Venezia, Dorsoduro 2137, 30123 Venice, Italy

c

CNR-Istituto per la Dinamica dei Processi Ambientali, Dorsoduro 2137, 30123 Venice, Italy

d

Institute of Marine Environment and Resources, 246 Da Nang Street, Haiphong City, Vietnam

a r t i c l e i n f o

Keywords:

Polychlorinated biphenyls

Sediments

Historical trends

Coastal lagoons

Central Vietnam

a b s t r a c t

PCBs were analysed in surficial sediments and selected sediment cores collected between 2002 and 2008

in Central Vietnam coastal lagoons The aim was to determine contamination levels and trends, and to evaluate the effects of anthropogenic pressures and natural events Samples were mostly fine-grained with low total PCB concentrations (0.367–44.7lg kg 1) Atmospheric transport and post depositional processes modify to some degree the fingerprint of PCB inputs to the environment favouring the predom-inance of 3, 4 and 5 chlorinated congeners The similarity of congener distributions in contemporary sur-ficial samples also suggests the presence of a unique source over the entire study area, probably connected to mobilisation and long range transports from land-based stocks The removal of consistent sediment layers is hypothesised based on repeated samplings of the same area Natural meteorological events (such as typhoons) are suspected to be responsible for these sediment losses

Ó2011 Elsevier Ltd All rights reserved

1 Introduction

Polychlorinated biphenyls (PCBs) are compounds with a wide

range of properties Commercially PCBs were marketed as mixtures

with a degree of chlorination from 21% to 68% The most commonly

used were those with a chlorine content between 42% and 54%, such

as Aroclor 1242 and 1254 (Barbalace, 2003) Starting from their first

commercialisation in 1929, these mixtures were used in many open

(e.g., as lubricants, softeners for plastics and glues, laminating

agents in paper production, impregnating agents, fire retardants,

printing inks, oils, paints, self-copying paper, cement plaster and

casting agents, insecticides, etc.) and closed systems (e.g., insulation

liquid in capacitors, insulation and cooling liquid in transformers,

hydraulic oils, heat exchangers), but their danger was soon

acknowledged and the production was first regulated and then

banned in the late 1970s (Erickson, 2001) Despite these efforts,

PCBs have become ubiquitous pollutants, due to many factors: (1)

the multiplicity of PCB sources and transport mechanisms; (2) the

fact that their use is still allowed in enclosed transformers and

capacitors, and (3) the presence of PCB in old ships as hydraulic

liquids and covers of electric cables (Hutzinger et al., 1974; Atlas

et al., 1986) In addition, landfills are now considered a major source,

due to their release from rubbish and wastes (Breivik et al., 2002) As

a consequence, PCBs still constitute a worldwide environmental problem, and the knowledge of their concentrations in the environ-ment is needed to understand present contamination levels and trends, to assess the risk, and to plan management strategies In this context, the assessment of sediment contamination is considered a

et al., 2001; Colombo et al., 2005; Kuzyk et al., 2005; Sundberg et al., 2005; Denton et al., 2006)

The information about PCB sources and levels in Vietnam is lim-ited However, it is likely that environmental levels have been exacerbated by the long periods of conflict and subsequent rapid

described the PCB distribution in cores and surficial samples from Tam Giang-Cau Hai (TG-CH) system, but nothing was known about the contamination of the other coastal lagoons in Central Vietnam

We aimed to fill this gap by assessing present PCB levels through the analysis of surficial samples Additionally, trends and control-ling factors were examined at five selected sites (four lagoons) through the study of sediment cores and the repeated sampling

of the same locations in different years

2 Study sites The studied lagoons are all located in the central part of the coun-try, between 11°N and 16°N (Fig 1), and belong to the coastal

0025-326X/$ - see front matter Ó 2011 Elsevier Ltd All rights reserved.

doi: 10.1016/j.marpolbul.2011.02.035

⇑ Corresponding author Address: Consiglio Nazionale delle Ricerche Istituto di

Scienze Marine, Sede di Bologna, Via Gobetti 101, 40129 Bologna, Italy Tel.: +39

051 6398864; fax: +39 051 6398940.

E-mail address: silvia.giuliani@bo.ismar.cnr.it (S Giuliani).

Marine Pollution Bulletin 62 (2011) 1013–1024

Marine Pollution Bulletin

j o u r n a l h o m e p a g e : w w w e l s e v i e r c o m / l o c a t e / m a r p o l b u l

provinces of Thua Thien Hué, Quang Nam, Quang Ngai, Binh Dinh,

Phu Yen, Khanh Hoa and Ninh Thuan They are diverse in scale, shape,

size, stability of inlets, water features and geological and

geomor-phological distribution (Cu, 1995), and can be rated into four

and Thi Nai, TN) These valuable and diverse ecosystems, important

as tourist attractions and useful for fishing and aquaculture

activi-ties, are key sites for the development of the Vietnamese economy

However, ongoing problems with anthropogenic pollution, such as

high concentrations of oil, nitrate and coliforms in water (Dieu,

2006; Thom, 2006) are believed to be deeply affecting the

environ-mental quality of these sites over time Although the concentrations

of persistent organic pollutants such as PCBs and PAHs (Frignani

et al., 2007; Giuliani et al., 2008) are believed to be low, these are also

expected to increase over time as economic development continues

3 Materials and methods

Sampling locations in Central Vietnam coastal lagoons are shown

inFig 1 Sediment cores were sampled multiple times from 2002 to

2008, using a manual piston corer (6 cm i.d.) that preserved undis-turbed sediment–water interfaces Collection dates and core lengths

sampling details of the TG-CH cores collected in 2002 and already

extruded and sectioned at intervals of 1–4 cm, with higher resolu-tion at the top Sediment slabs, cleaned at the edges to avoid the effects of smearing, were then put in polyethylene vessels and stored

in a refrigerator at 0 °C until arrival at the lab Afterwards, they were conserved at 18 °C until analysis

Grain size analyses were carried out by wet sieving, to

frac-tions were determined with a X-ray Micrometric SediGraph Organic carbon (OC) content was obtained through a CHN ana-lyser after elimination of carbonates by treatment with HCl in

a silver capsule

were obtained by gamma counting of dry samples in standard vessels of suitable geometries Radiotracer analyses were described

Collection dates and results of the analysed parameters (core length, porosity,% content of fines, i.e silt plus clay, 210 Pb activity, total and fine-normalised PCB concentrations,% contribution of each PCB congener class, from 1 to 10 Cl substitutes) for all surficial samples of the cores collected from 2002 and 2008.

Lagoon Collection

date

Core length (cm)

Porosity Fines

(%) a

activity (Bq kg 1 )

PCBs (lg

kg 1 )

Norm PCBs (lg kg 1 )

CB-1 (%)

CB-2 (%)

CB-3 (%)

CB-4 (%)

CB-5 (%)

CB-6 (%)

CB-7 (%)

CB-8 (%)

CB-9 (%)

CB-10 (%)

n.d = Not detected.

a Grain sizes are calculated considering only the fraction <2 mm.

b The sediment is entirely sandy; a silt plus clay value of 1% was assumed for normalisation.

c Samples from OL, CR and DN contain 21.5%, 6.3% and 6.3% of shell fragments and small gravels, respectively; normalising with respect to the real content of fines would give PCB concentrations of 5.52, 10.6 and 27.1lg kg 1

PCBs were extracted from 1 to 3.5 ± 0.01 g of lyophilised

Applied Separations, LabService Analytica) using

dichlorometh-ane/acetone (1:1 v/v) in presence of anhydrous sodium sulphate,

diatomaceous earth and activated metallic copper The clean-up

System (Di-oxin Prep, Fluid Management System, Inc., LabService Analytica)

using a column of neutral silica and elution with n-hexane and

then n-hexane/dichloromethane (1:1 v/v) PCBs were determined

by High Resolution Gas Chromatography-Low Resolution Mass

Spectrometry (HRGC-LRMS) with a Hewlett Packard Model 6890

Gas-chromatograph coupled with a Hewlett Packard model 5973

Mass Selective Detector (mass analyzer: quadrupole) according

toMoret et al (2005) Analytical details are fully described in

Mor-et Mor-et al (2001)andPiazza et al (2009) High-Resolution Mass

Spec-trometry (MAT 95 XP, Thermo Finnigan, Bremen, Germany) was

also used for the identification of PCB congeners For the

Laboratories, Andover, Massachusetts, USA) were added to the

10-CB Crude concentration values were corrected with

congener-spe-cific instrumental response factors obtained by measuring four PCB

standard solutions (C-CS01, C-CS02, C-CS03 and C-CS05 by

Accu-Standard, Inc., New Haven, USA), for a total of 122 congeners Only

those peaks with heights equivalent to at least three times the

background value were considered for identification and

subse-quent quantification GC–MS detection limits were 1 pg for all

congeners, whereas at 0.6 pg level, 80% of them were detectable

Eighty-three chromatographic peaks, representing 101 PCBs

cong-eners (14 as double and 2 as triple peaks), were detected: two

2-CBs (P2-CBs 12 and 15), nine 3-2-CBs (P2-CBs 18, 17, 24 + 27, 16 + 32,

26, 25, 28 + 31, 20 + 33 and 22), 14 4-CBs (PCBs 45, 46, 52, 49,

47 + 48, 44, 42 + 59, 41 + 64 + 71, 40, 74, 70, 66, 56 + 60 and 77),

16 5-CBs (PCBs 93 + 95, 91, 92, 84 + 90 + 101, 99, 97, 87 + 115,

83, 85, 110, 82, 109, 119, 123, 118 and 105), 18 6-CBs (PCBs 136,

151, 135 + 144, 147, 149, 134, 131, 146, 153, 132, 141, 137,

138 + 164, 128 + 167, 156, 158, 129 and 157), 16 7-CBs (PCBs

179, 176, 187, 178, 183, 174, 185, 177, 171, 172, 173, 180, 193,

191, 170 + 190 and 189s), six 8-CBs (PCBs 196 + 203, 197, 201,

195, 194 and 205), one 9-CB (PCB 207) and PCB209 (the sole

10-CB) All solvents used were pesticide grade (Labscan Ltd., Dublin,

Ireland) Accuracy was checked at the time of analysis using a

cer-tified standard sediment (NIST, Standard Reference Material

1941b) The observed concentrations fell within the 95%

confiden-tial intervals reported in the certificate of analysis Reproducibility,

based on six analyses of the same certified standard, was 4% for the

sum of congeners, and between 0% and 9% for each individual

congener

All surficial samples (0–1 cm) were analysed, together with

se-lected samples from cores colse-lected in 2004 and 2007 from TG-CH

Lagoon (just core 10 in 2004), in 2005 from TN and in 2008 from

LC, TN and OL PCB concentrations of the TG-CH 2002 cores are

4 Results

4.1 Surficial samples

Porosity values of surficial samples collected in the period

2002–2008 range from 0.45 to 0.94 and remain quite constant

between repeated samplings The highest differences are observed

at CR, DN, and NN, probably linked to grain size variations (see below)

Most surficial samples are fine grained (clayey silt or silty clay) with sands prevailing only at TG, TN and DN (Table 1) When com-paring the samples over time, a shift to a coarser composition is observed at CR and NN, with a doubling in the contribution of sand from 2005 to 2008, and the presence of shells and small gravels up

to 6.3% in 2008 The grain size composition remains unvaried at LC and OL, while there is a decrease of the sandy fraction in the 2008 samples from TN and DN The samples from TG-CH 10 showed an increase in the sandy fraction from 2002 to 2004 and a subsequent decrease to values finer than the original sample in the period 2004–2007 (Table 1)

consis-tently higher in the first samplings A general decreasing trend is observed from the northernmost locations (TG-CH and LC) towards

being one of its decay products, in the Thua Thien-Hué region (where the two lagoons are located) The natural enrichment of

Quang Nam deposit near Hué (Kušnír, 2000)

4.1.2 PCB concentrations and compositions Total PCB concentrations in surficial samples from the 2002–

the most recent samples are higher that the previous ones (from 2

to 3 times, as observed at OL and DN, respectively) or roughly con-stant (LC) In contrast, TG-CH 02, TG-CH 10 and TN present total PCB concentrations in the earlier samples (2002 and 2005, respec-tively) that are considerably higher than subsequent measurements (up to 13 times in the case of TN) The decreasing trend continues at TG-CH 10 from 2004 to 2007, while in the same period a slight

In order to compare surficial concentrations without the effects

of sediment grain size composition, PCB values were normalised to the content of silt plus clay, thus determining what PCB concentra-tions would be expected if the sediments were entirely fine grained (Table 1) This effect is minimal in the majority of samples classified as clayey silt or silty clay but becomes more apparent with the increasing content of sand This is the case for TG in

2005 and NN, TN, TT and DN in 2008, and gets to its greatest extent

in TN and DN in 2005, where the normalisation enhances PCB con-centrations up to two orders of magnitude due to the completely sandy composition of the surficial sediments (Table 1)

The PCB homologue composition of 2002–2008 surficial

concentration In general, 3, 4 and 5-CBs together account for the greatest portion of PCBs, averaging 91% and being less than 70%

in only two samples: the first, NN (0–1) from 2005, shows a signif-icant contribution of 6-CBs, and a smaller content of 4-CBs, whereas the other, TN (0–1) from 2005, has practically no 3 and 4-CBs and is composed almost exclusively of heavier 5, 6 and 7-CBs As for the TG-CH cores, the 2002 samples contain almost entirely 3, 4 and 5-CBs, whereas the later samples showed a slight increasing presence of both lighter (2-CBs) and heavier (6, 7-CBs) congeners Additionally, in the 2004 TG-CH 10 core, some 8, 9, and 10-CBs were detected

4.2 Sediment cores

Results from cores collected in 2002 from TG-CH and in 2004–

2005 from the eight minor lagoons have been already described by Frignani et al (2007) and Giuliani et al (2008), respectively

However, they are here briefly summarized inFig 2, together with

new data from the 2004, 2007 and 2008 campaigns

Core sample porosity ranges from 0.51 (TG-CH 10 in 2004) to

0.78 (OL in 2005) and is generally higher at surface, follows a

decreasing trend in the intermediate layers, then remains quite

constant down to the core bottom TG-CH 02 in 2007 and TN in

2005 differ slightly from this pattern, in that they have deep sedi-ment values that are higher than surface ones In addition, porosi-ties of TN in 2008 constantly increase with depth and the value measured at surface is among the lowest Despite these limited

Fig 2 Depth distributions of porosity, contents of sand and fines,210Pb and total PCBs in sediment cores collected from TG-CH, LC, TN and OL coastal lagoons during different sampling campaigns (2002–2008) PCB homologue patterns at selected depths are also shown.

S Giuliani et al / Marine Pollution Bulletin 62 (2011) 1013–1024 1017

differences, in general profiles are quite similar between repeated

samplings

The prevalence of fine sediments shown by surficial values (see

above) is also confirmed by core profiles: the sand content is

signif-icantly below 50% in most cores, with fine particles accounting for

more than 75% of the total The sole exception is TN where both

samples have surface and subsurface sand percentages above

50% Surface sediments have already been discussed in the relative

section (see above), but grain size profiles show that the sandy

fraction remains significant down to 20 and 30 cm depth (in the

2005 and 2008 core, respectively), below which the content of fine

grains is similar to those of the other lagoons

evidence of post-depositional reworking, such as bioturbation

pro-cesses Unsupported background values (i.e those produced by

depths, from 5 cm (TN in 2008) to 30 cm (LC and TG-CH 02 in

core chronologies is weakened in these environments because of

4.2.2 Profiles of PCB concentrations and compositions

0.5 cm depth) Trends are quite constant or increase significantly near surface (this is the case for TG-CH 02 and 10 in 2002 and

TN in 2005) As for TG-CH Lagoon, the 2002 values are generally higher than the corresponding layers from the later campaigns Congener composition is variable not only between sites, but also within core repetitions and along the same sedimentary re-cord While the 3, 4 and 5-CBs (with a minor contribution from 6-CB) account for the majority of PCBs as observed in surficial samples, their relative importance differs between layers (Fig 2) The presence of other PCB congeners (i.e 2-CBs and from 7 to 10-CBs) occurs only in TG-CH 02 and 10 in 2004 In general, the comparison of repeated samples shows that OL remains constant

Fig 3 Worldwide distribution of average PCB levels in surficial sediments compared to Central Vietnam coastal lagoons Values are grouped into three major sections: contaminated sites ( Van Bavel et al., 1995; Camacho-Ibar and McEvoy, 1996; Sundberg et al., 2005 ), Coastal and industrial areas ( Tolosa et al., 1997; Khim et al., 1999; Laane

et al., 1999; Müller et al., 1999; Nhan et al., 1999, 2001; Pettersen et al., 1999; Frignani et al., 2001; Lee et al., 2001; Ma et al., 2001; Santschi et al., 2001; Barakat et al., 2002; Fillman et al., 2002; Bertolotto et al., 2004; Frignani et al., 2004; Saponizhnikova et al., 2004; Colombo et al., 2005; Kuzyk et al., 2005; Denton et al., 2006; Hung et al., 2006 ) and Lagoons ( Frignani et al., 2001, 2004; Konat and Kowalewska, 2001; Menone et al., 2001; Moret et al., 2001; Fillman et al., 2002; Secco et al., 2005 ) Maximum and minimum range bar for Central Vietnam lagoons is also shown International sediment quality guidelines are indicated as horizontal continuous (ERM) or dotted (ERL) lines.

from 2005 to 2008 (with the predominance of 4-CBs in both

sam-plings), while LC and TN display a shift over time to lower

Cl-substituted congeners (from 5 and 6-CBs to 4-CBs) A similar

temporal pattern can also be observed for TG-CH 10, where the

proportion of 3-CBs increases from 2002 to 2007 TG-CH 02 is

com-pletely different, with 3-CBs present only in 2002 and at depth in

2007 In this latter core, 5-CBs are predominant at the surface

and at 6–8 cm depth

5 Discussion

5.1 PCBs in Vietnamese coastal lagoons and their potential threat to

the environment

The average surficial sediment concentration found in this

of low impacted coastal and lagoon areas reported in the literature

for globally distributed environments (Fig 3), and is up to three

or-ders of magnitude lower than those observed in the most

contam-inated sites The highest value measured in the TN Lagoon in 2005

of most sediment quality guidelines used around the world, above

which adverse effects to biota are expected (e.g., NOAA ERM

present PCB sediment concentrations of Central Vietnam coastal

lagoons do not constitute a threat to the environment Even in case

of resuspension of deep sediment layers (which can happen during

dredging, floods, or storms), the risk of contamination is absent

since downcore PCB values, in most cases, are lower than at the

surface (Fig 2) Nevertheless, increasing values found in some of

might be an indicator of enhanced recent human pressure that

sug-gests the need for a monitoring plan to prevent the possible

dangerous worsening in the near future

5.2 PCBs sources, transport mechanisms and natural degradation

processes

In order to evaluate the origin of the PCB patterns found in

sur-ficial samples from all sampling campaigns, the homologue

com-position was compared with that of the principal commercial

mixtures Aroclor 1016, 1242, 1254, 1260 and 1268 were

consid-ered, and their compositions were normalized with respect to

the list of congeners found in samples Autoscaled data were

processed by a cluster analysis (Ward’s method as aggregative

clustering, using the Euclidean distance as similarity measure),

smal-ler one, are well distinguished, with each essentially containing samples collected in the same period Starting from the right of Fig 4, the first group (‘‘a’’, with the highest similarity) is composed

of all 2008 samples from the minor lagoons plus those from TG-GH

in 2002 and is not associated with any of the commercial Aroclor

(‘‘b’’, Fig 4) is composed of the 2004–2005 samples of the minor lagoons (with the exception of TN) and the 2007 repetitions at TG-CH, with the predominance of 4- and 5-CBs, resembling Aroclor

1248 and 1254, respectively (EPA, 2008) Moreover, the 2004

strictly linked to any of the commercial mixtures that we consid-ered The finding that contemporary samples present similar PCB compositions suggests a unique source type over the entire study area which should be connectable to mobilisation processes from land-based stocks, such as evaporation from products and contam-inated soils, spills or leakages from landfills or incinerators, or im-proper disposal of equipments still containing PCBs These processes may be further enhanced in tropical and subtropical re-gions by the prevailing climatic conditions These latter favour the remobilisation and volatilisation of PCBs into the atmosphere (Iwata et al., 1994) also from remote locations, as confirmed by the fact that PCB fluxes through atmospheric currents are the larg-est reaching marine environments (Zarfl and Matthies, 2010) In-deed, the low PCB values in the studied sediments could imply their atmospheric transport from very distant zone with respect

to more localised point sources, as happens, for example, in much more polluted areas where local sources are predominant over

Sommerfreund et al., 2010) However, the discrimination between local and remote sources is impeded by the lack of information rel-ative to the composition of PCB mixtures present over the territory

In addition, the differences observed among repeated samples of the same area are most likely linked to structural modifications

of PCB compositions due to post-depositional processes, instead

of a change in sources Indeed, there is no evidence that could jus-tify such source variation, with the exception of the TN Lagoon where the 2005 surficial sample is by itself on the left side (‘‘d’’)

ofFig 4and is somewhat related to Aroclor 1260 This is not sur-prising since the TN-05 congener composition presents practically

no 3- or 4-CBs, as it is composed almost exclusively of the heavier 5-, 6-, and 7-CBs as in the above mentioned commercial mixture (EPA, 2008) This particular sample is completely different from the lower layers and the samples collected in 2008 at the same site, with both groups resembling the general 3-, 4-, and 5-CB

predom-Ward`s method - Euclidean distances

0 5 10 15 20

1268 1260 TN-05 1242 1016 10-04 02-04 1254 10-07 NN-05 1248 CR-05 OL-05 DN-05 TG-05 NM-05 02-07 LC-04 NM-08 NN-08 DN-08 CR-08 LC-08 OL-08 TN-08 10-02 02-02

a b

c

d

Fig 4 Cluster analysis (Ward’s method as aggregative clustering, using the Euclidean distance) comparing the composition of PCBs in surficial samples from Central Vietnam coastal lagoons with the most common Aroclor commercial mixtures Main clusters (a–d) are indicated.

S Giuliani et al / Marine Pollution Bulletin 62 (2011) 1013–1024 1019

inant composition Additionally, this sample shows the highest PCB

concentration of the entire dataset and is the only one in the

pres-ence of a relatively high PCB concentration in a completely sandy

sediment (not suited to interact with and store contaminants) is

very unusual These evidences, together with the contemporary

observation of a different congener distribution, account for the

ef-fect of external point sources, most likely linked to the activities

carried out in the nearby Phuong Mai peninsula in the framework

of the Nhon Hoi Economic Zone (NEZ) development plan Indeed,

City in the Binh Dinh province where the TN Lagoon is located It was completed in 2010 and includes residential areas, an industrial

NEZ is connected to Quy Nhon’s city centre by the Thi Nai Bridge that was under construction in 2005, just when the TN core was

Fig 5.210Pb activity-depth profiles from repeated samplings in the TG-CH, LC, TN and OL lagoons Two subsequent campaigns are compared (2002–2004 and 2004–2007 for TG-CH 02 and TG-CH-10, 2004–2008 for LC and 2005–2008 for TN and OL), the values of the older ones being corrected for 210 Pb natural decay Depths relative to one sample are shifted downward (as specified in the relative legend) to provide a satisfying superposition of the two profiles Grey areas identify the hypothesized thicknesses of removed sediments.

collected As none of the analysed sediment layers of the core

collected in 2008 present either the PCB levels measured at the

surface or its congener composition in 2005 suggests a recent

recovery which can be attributed to both degradation and removal

processes (see the following section), or lateral inhomogeneity

If the changes in PCB homologue profiles in surficial samples

can be explained based on sediment dynamics, it is then difficult

to explain the observed differences downcore However, in most

cases the composition remains constant, or nearly constant,

throughout the core (e.g., LC, 2008; TG-CH 02, 2002; TG-CH 10,

2007; TN, 2008; OL 2008), but at certain places the PCB pattern

has changed over time However, at some locations the PCB

composition is only slightly different (e.g., TG-CH 10, 2002), and

in almost all cases different sediment levels were analysed in cores

taken in different years

As for the differences among repeated samplings of the other

la-goons, the congener patterns define a general trend towards the

enrichment of lower Cl-substituted classes (3- and 4-CBs) in the

control the behaviour of PCBs in the environment favour the

mobility of the more volatile congeners, i.e those with lower

chlo-rination degree (Gevao et al., 1998), so that sediments from several

marine coastal locations generally present congener patterns in

which 5-, 6-, 7-, or 8-CBs contribute to a greater proportion to

the total (Khim et al., 1999) In spite of that, not only lighter

cong-eners dominate in sediments of Vietnamese lagoons but their

rel-evance also increases with time This situation might not be so

5-CB dominated patterns in Mexican and Moroccan aquatic

sedi-ments Coastal marine settings may be different due to the use of

lower chlorinated PCB mixtures in these countries, or, most likely,

to easier long range transport and incorporation of less chlorinated

PCBs in these sediments Indeed, PCB fractionation along

horizon-tal gradients is driven by the atmospheric persistence of individual

congeners that is higher for the lighter, less chlorinated PCBs

(Schuster et al., 2010) Therefore the predominance of 3- and

4-CBs in the studied sediments can be explained by the prevalence

of atmospheric-derived inputs from distant anthropogenic

loca-tions, following predominant wind directions These blow from

southwest during the hot-rainy season (May–October) and from

et al., 2010), indicating the urbanized areas of Southern Vietnam

and South-western China as potential source areas for PCB

contamination

Finally, biologically mediated degradation processes can also

modify the PCB composition under both aerobic and anaerobic

conditions The former mineralises the pollutants to its constituent

chlorine atoms from highly chlorinated PCBs (Lake et al., 1992;

Borja et al., 2005), namely the flanked ones followed by those in

meta positions (Karcher et al., 2007) In sewage sludges and under

strict methanogenic conditions, PCB removals have been estimated

as about 40% for all congeners, while strong aerating actions

en-hanced the removal of the lightest PCBs up to 100% (Patureau

and Trably, 2006) However, biotic PCB degradation may be slowed

down in natural environments due to binding of PCBs to sediment

surfaces (Strand, 1990) and is also strongly limited by PCB

bioavail-ability (Patureau and Trably, 2006) In synthesis, anaerobic

pro-cesses active in our sediment cores might have increased the

presence of lower chlorinated PCBs (3- and 4-CBs of the minor

la-goons and 2-CBs in TG-CH in 2004) in the 2004–2008 samples,

while mineralisation might explain the disappearance of 2-CBs

from TG-CH sediments in 2007 However, a correct assessment of

the importance of these processes requires further dedicated

research

5.3 Influence of natural and anthropogenic events as revealed by repeated sedimentary sequences

The research strategy that combines the study of temporal re-cords from cores and repeated sampling is suited for the identifica-tion of processes that control and modify the sedimentary sequence and the presence of contaminants in the environment

In this case we can rely on both records and can observe some interesting features that could be in accordance with the effects

of natural and anthropogenic events

shows that trends are similar, just shifted slightly in depth (Fig 2)

of the profiles becomes quite good when they are decay corrected and transposed downward (Fig 5) As an example, a highly

accounts for the representativeness of the second and third

or unable to disguise the basic information On the other hand, it indicates that there are major changes over time, interesting

of TG-CH 02 almost ovelaps that obtained in 2002 (once corrected

the two samplings) if it is shifted of about 7 cm downward This re-sults by comparing the intercept (a) and the angular coefficients (m)

of the equations (in the general form y = mx + a) representing the

are almost parallel and the shift is given by the difference (7) be-tween the intercepts (a): 19.6 and 12.6 cm for 2002 and 2004

sed-iment layer of at least 7 cm (when further sedsed-imentation is not con-sidered) has been removed from the site of core TG-CH 02 at some point between 2002 and 2004, and was displaced elsewhere Similar conclusions can be drawn for the first two samples of the other la-goons, with varying thicknesses of the removed layer (from 3 cm

con-firmation, the shifts in the second samples, attributed to sediment losses, produce a certain superposition of observed trends when ap-plied to other sedimentary parameters (i.e porosity and grain size

dis-crepancies can be noticed in sand profiles, in particular at TN, due

y = -0.0826x + 12.60

R 2 = 0.9853

y = -0.0637x + 19.60

R 2 = 0.8166

0

5

10

15

20

Bq kg -1

decay corrected from 2002 measured in 2004 TG-CH 02

Fig 6 Linear regression of the upper portion of 2002 and 2004 210 Pb activity-depth profiles at TG-CH 02 Assuming the similarity of the angular coefficients, the lines almost superimpose themselves if they are shifted of 19.6–12.6 = 7 cm.

S Giuliani et al / Marine Pollution Bulletin 62 (2011) 1013–1024 1021

Fig 7 Profiles of porosity and sand content modified according to the depth shifts shown in Fig 5 As for TG-CH Lagoon, only the first two sampling campaigns (2002 and 2004) are considered here.