Food Chemistry 151 (2014) 7278

Contents lists available at ScienceDirect

Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem

Analytical Methods

Development, validation and accreditation of a method for the

determination of Pb, Cd, Cu and As in seafood and sh feed
samples
A.K. Psoma, I.N. Pasias, N.I. Rousis, K.A. Barkonikos, N.S. Thomaidis
Laboratory of Analytical Chemistry, Department of Chemistry, National and Kapodistrian University of Athens, Panepistimiopolis Zografou, 15771 Athens, Greece

a r t i c l e i n f o a b s t r a c t

Article history: A rapid, sensitive, accurate and precise method for the determination of Pb, Cd, As and Cu in seafood and
Received 19 July 2013 sh feed samples by Simultaneous Electrothermal Atomic Absorption Spectrometry was developed in
Received in revised form 12 October 2013 regard to Council Directive 333/2007EC and ISO/IEC 17025 (2005). Different approaches were investi-
Accepted 10 November 2013
gated in order to shorten the analysis time, always taking into account the sensitivity. For method vali-
Available online 20 November 2013
dation, precision (repeatability and reproducibility) and accuracy by addition recovery tests have been
assessed as performance criteria. The expanded uncertainties based on the Eurachem/Citac Guidelines
Keywords:
were calculated. The method was accredited by the Hellenic Accreditation System and it was applied
Seafood
Fish feed
for an 8 years study in seafood (n = 202) and sh feeds (n = 275) from the Greek market. The annual
Eight-year study and seasonal variation of the elemental content and correlation among the elemental content in sh feeds
Heavy metals and metalloids and the respective sh samples were also accomplished.
Uncertainty calculation 2013 Elsevier Ltd. All rights reserved.
Seasonal variation

1. Introduction criteria, by which an authoritative body provides formal recogni-

During the last few years, seafood consumption has undergone
major changes. The rapid growth in aquaculture production has
made the sector important to the economy of many developing
countries and, in the case of some traded aquatic products; the sec-
tor has become either an important source of supply or the main
supplier. On the contrary, the increased concern for public health
has obligated the European Commission to set maximum permis-
sible levels for major toxic elements in food. Due to the availability
and variety of different analytical methods for the determination of
many residues in food samples, it is necessary to ensure the quality
and comparability of the analytical results generated by laborato-
ries approved for ofcial control. For this reason, there is a growing
global interest in performing analytical methods which can pro-
vide accurate and precise measurements with low detection limits
and costs in short analysis times. This should be achieved by using
quality assurance systems and specically by applying methods
validated according to common procedures and performance crite-
ria (European Commission Decision, 2002; European Commission
Directive, 2002; European Commission, 2006, 2007, 2008a). In
the elds of chemical analysis, accreditation covered by ISO/IEC
17025 (2005) is the procedure, based on internationally-agreed
Corresponding author. Tel.: +30 210 7274317; fax: +30 210 7274750.
E-mail address: ntho@chem.uoa.gr (N.S. Thomaidis).
tion that a body or a person is in a position to carry out specic
tasks with impartiality and competence. ISO/IEC 17025 (2005)
species the general requirements for the competence to carry
out tests and/or calibrations, including sampling. It covers testing
and calibration performed using standard methods, non-standard
methods, and laboratory-developed methods. ISO/IEC 17025
(2005) is for use by laboratories in developing their management
systems for quality, administrative and technical operations (ISO/
IEC 17025 (2005)). The Hellenic Accreditation System (ESYD) has
been appointed as the National Accreditation Body of Greece
according to the requirements of Article 4 of the Regulation (EC)
No. 765/2008, according to which each Member State shall appoint
a single national accreditation body (European Commission,
2008b).
Numerous studies have been carried out on metal accumulation
in different seafood and sh feed samples (Low, Zain, & Radzi Abas,
2012; Maule, Gannam, & Davis, 2007; Mendil, Demirci, Tuzen, &
Soylak, 2010; Tuzen, 2009; Trkmen, Trkmen, Tepe, Ates, &
Gkkus, 2008; Trkmen, Trkmen, Tepe, & eki, 2010). In these
studies different analytical techniques, including inductively cou-
pled plasma-atomic emission spectrometry (ICP-AES) (Trkmen
et al., 2008), inductively coupled plasma mass spectrometry (ICP-
MS) (Low et al., 2012), ame (FAAS) and graphite furnace (GFAAS)
atomic absorption spectrometry (Mendil et al., 2010) were used for
the determination of trace elements after matrix destruction based

0308-8146/$ - see front matter 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.foodchem.2013.11.045
 A.K. Psoma et al. / Food Chemistry 151 (2014) 7278
on oxidation with concentrated acids. However, to the best of our
knowledge, no accredited methods have been reported in the liter-
ature concerning the simultaneous determination of Cd, Pb, Cu and
As in seafood and sh feed samples. Furthermore, there are few
studies concerning the seasonal variation of these elements in sea-
food samples (Ekpo & Ibok, 1999; Mendil et al., 2010; Obasohan &
Eguavoen, 2008), but the collected data sets provide conclusions
only about the correlation between the seasonal variability and
the elemental annual accumulation.
In this study, a method for the simultaneous determination of
Pb, Cd, As and Cu in seafood and sh feed samples by Simultaneous
Electrothermal Atomic Absorption Spectrometry (SIMAAS) was
developed in regard to Council Directive 333/2007 EC and ISO/
IEC 17025 (2005) was accomplished. For this reason, different ap-
proaches, including the single-elemental, the two elemental deter-
mination (PbCd and CuAs) and the simultaneous determination
of all the four analytes, were investigated in order to shorten the
analysis time. The developed method was accredited by the Hel-
lenic Accreditation System and it is now available for the ofcial
control of Pb, Cd, As and Cu in seafood and sh feed samples. Fur-
thermore in this study, an overall seasonal variation of metal accu-
mulation for over 8 years for the sea breams and sea basses
collected from aquaculture centres in Greece and the seasonal var-
iation of the metal content in sh feeds, collected together with the
sea bass and sea bream samples from the same aquaculture centres
during 20052013, was performed. Finally, the metal content of 17
additional seafood species, purchased from the Greek market in
different years was determined for the rst time. Due to the vast
number of collected data sets, reliable conclusions about the corre-
lation between temporal variability and element concentration can
be deduced.
2. Materials and methods
2.1. Instrumentation
A Perkin-Elmer SIMAA-6000 simultaneous multielement atom-
ic absorption spectrometer with a longitudinal Zeeman-effect
background correction system, equipped with a transversely-
heated graphite atomizer (THGA), and an AS-72 autosampler were
used. Electrodeless discharge lamps (EDL) were used for As
(193.7 nm and slit 0.7 nm), Pb (247.6 nm and slit 0.7 nm), and
Cd (228.8 nm and slit 0.7 nm) respectively, and a hollow cathode
lamp was used for Cu (324.8 nm and slit 0.7 nm) and they were
operated according to the conditions recommended by the manu-
facturer. Atomic signals were measured in peak area mode. It
should be pointed out that the experiments were carried out
using stabilized temperature platform furnace (STPF) conditions,
including Zeeman-effect background correction. The digestions
were performed on a MARS X-Press microwave oven (CEM
Corporation).
2.2. Reagents, analytical solutions and samples
Ultra pure water (18.2 MX cm at 25 C) from a Millipore Direct-
Q-UV water purier (Millipore) obtained using a Milli-Q system
(Millipore). All solutions were prepared with Suprapur grade nitric
acid (Merck, Darmstadt, Germany). The argon was of 99.999%
purity.
The mixed standard solution containing 50 lgAs L1, 50 lgCu L1,
50 lgPb L1 and 5 lgCd L1, were daily prepared by appropriate
dilution of 1000 mg L1 As, Cu, Pb and Cd stock standard solutions
and acidied with nitric acid to 1% (v/v). The working solutions
were prepared automatically by the autosampler from the mixed
standard solution.
73
The palladium/magnesium modier solution was prepared by
appropriate dilution of 10 g/L Pd (Merck) in HNO3 and 1 g/L Mg
in 1% (v/v) HNO3 (High Purity Standards, Charleston, SC). The Zr,
Ir and NH4H2PO4 solutions were prepared by appropriate dilution
of 1000 5 lg mL1 ZrOCl22H2O in 0.5% HNO3 (High Purity Stan-
dards), 1000 lg mL1 of Ir in 2% HCl (High-Purity Standards) and
100 g/L NH4H2PO4 in 0.05% HNO3 (High-Purity Standards) with ul-
tra pure water. Likewise, HNO3 65%, Suprapur (Merck) was used
for digestion of the samples.
Nineteen different seafood species (totally 202 samples) and
275 sh feed samples were obtained from local markets in Athens,
Greece and different aquaculture farms from Aegean and Ionian
Sea in Greece (from 2005 to 2013). Samples were milled and
homogenised in a food processor as they arrived in the laboratory
with a stainless steel cutter to obtain a representative homogenous
subsample. For sh samples, only the edible muscle was homoge-
nised whereas, for shrimps, and cephalopods, the whole body was
homogenised. Finally, only the soft tissue of the mussels was
homogenised, after rejection of the shell. The homogenised sam-
ples were stored at 18 C till analysis.
2.3. Analytical procedure
All glassware and polypropylene bottles were cleaned with
acidied ultra pure water. 0.5000 g (wet weight) of the homoge-
nised seafood or 0.2500 g (wet weight) of the homogenised sh
feed samples were weighted into the Teon vessels and 5 mL of
65% HNO3 were added. The samples were digested with the MARS
X-Press (CEM Corporation) microwave oven with a preselected
programme (rst stage: 1600 W, 2 min ramp time, 165 C, 0 min
hold time; second stage: 1600 W, 3 min ramp time, 175 C, 5 min
hold time) and then diluted to a nal volume of 20 mL with ultra-
pure water.
The pyrolysis and atomization temperature curves were ob-
tained by spiking 20 lL of a digested sample solution and 20 lL
of the mixed standard solution in order to investigate the electro-
thermal behaviour of those elements. For this purpose, the ele-
ments were tested separately, in pairs (PbCd and AsCu), or in
an all elements mode in the presence of the different chemical
modiers.
Quantication was performed with calibration curves con-
structed from integrated absorbance measurements and prepared
by spiking 20 lL of different mixed aqueous or matrix standard
solutions of Pb, Cd, As and Cu. Standard addition calibration curves
were also constructed by spiking different standard solutions in
both seafood and sh feed samples in order to exploit the possibil-
ity for matrix interference. The instrumental limits of detection
(LOD (lg L1)) and of quantication (LOQ) were calculated from
the equations LODinst = 3.3  SBL/b and LOQinst = 10  SBL/b, respec-
tively, where SBL was the standard deviation of ten blank determi-
nations and b the slope of the curve. The LODs of the method were
calculated from the equation LODmeth (ng/g) = LODinstr (ng/
mL)  Vdilution (mL)  m (g)1. According to Commission Regulation
333/2007/EC, the LODsmeth and LOQsmeth for Cd and Pb should be
less than one tenth and less than one fth, respectively, of the max-
imum level in Regulation (EC) No. 1881/2006.
Precision under repeatability and reproducibility conditions
were also estimated from the calibration curves using sea bass
and sh feed samples with various concentration-level matrix
spiked solutions. As methods precision performance criteria, the
HORRATr were used, meaning the observed relative standard devi-
ation (%RSDr) under repeatability conditions divided by the RSDr
value estimated from the Horwitz equation (Thompson, 2000)
using the assumption r = 0.66R, as well as the HORRATR values,
meaning the observed RSDR value under reproducibility divided
by the RSDR value calculated from the Horwitz equation. According
74 A.K. Psoma et al. / Food Chemistry 151 (2014) 7278

Table 1
Optimal compromised conditions and the calculated analytical characteristics for all the three different approaches in seafood matrix solutions using Pd as chemical modier.

Analytical characteristics Simultaneous determination of Single element determination of

CdPbAsCu CdPb AsCu Cd Pb As Cu
Tpyr (C) 400 500 1100 500 700 1100 1100
Tatom (C) 1900 1500 2000 1400 1500 2000 1900
mo (pg) 2.880.044.019.6 1.935.2 41.924.4 1.6 48.9 40.2 21.0
Slope 0.03110.00110.00200.0045 0.04520.0025 0.00210.0036 0.0541 0.0018 0.0022 0.0042
LOD (lg/L) 0.0816.51.27 0.0130.44 0.271.4 0.030 0.38 0.065 0.70

to Commission Regulation 333/2007/EC, the HORRATr and HOR-
RATR values should be less than two. Furthermore, the accuracy
of measurements was also assessed through recovery experiments
of the addition of known amounts of the element to sea bass and
sh feed samples. Recovery data are only acceptable when they
are within 20% of the target value (European Commission Deci-
sion, 2002; European Commission Directive, 2002; European Com-
mission, 2006, 2007, 2008a).
Internal quality control charts (IQCs) were also constructed in
order to monitor whether results are reliable enough to be re-
leased. The objective of IQCs is the elongation of method valida-
tion: continuously checking the accuracy of analytical data
obtained from day to day in the laboratory. The analytical system
is under control if no more than 5% of the measured values exceed
the warning limits and none of them the action or control limits
(Taverniers, De Loose, & Van Bockstaele, 2004).
The uncertainty of the method was also calculated based on the
Eurachem/Citac Guidelines. For most purposes in analytical chem-
istry, an expanded uncertainty U should be used. The expanded
uncertainty provides an interval within which the value of the ana-
lyte concentration is believed to lie within a higher level of con-
dence. U is obtained by multiplying uc(y), the combined standard
uncertainty, by a coverage factor k. The combined uncertainty
uc(y) was calculated from the summary squared of several inde-
pendent parameters as: (a) the mass uncertainty; (b) the stock
standard solutions; (c) the volume uncertainty; (d) the calibration
uncertainty; (e) the bias uncertainty; and (f) the random errors
uncertainty, following the rules of Eurachem/Citac Guidelines.
Methods suitable for ofcial control must produce results with
standard measurement uncertainties less than the maximum stan-
dard measurement uncertainty calculated using the formula below
(European Commission, 2007):
q high calculated sensitivity in short analyses time. The simulta-
Uf LOD=22 aC 2
where:
Uf is the maximum standard measurement uncertainty (lg/kg).
LOD is the limit of detection of the method (lg/kg).
C is the concentration of interest (lg/kg).
a is a numeric factor to be used depending on the value of C.
3. Results and discussion
3.1. Comparison of single and simultaneous elemental determination
modied by Kalantzis, Rousis, Pasias, Thomaidis, and Piperaki
(2012).
Pyrolysis and atomization curves in both standard and matrix
solutions were established to determine the optimum pyrolysis
and atomization temperatures for each modier, used at the opti-
mal mass (Suppl. Fig. 1). When a temperature programme for
simultaneous multi-element ETAAS analysis is to be developed,
compromise furnace temperatures have to be considered related
to the group of analytes to be determined. Hence, the compromised
parameters are completely different than individually optimised
single-element conditions. The maximum pyrolysis temperature
(Tpyr) that could be applied is imposed by the most volatile ele-
ment, which in this case was Cd and the atomization temperature
(Tat) is dictated by the most refractory element, Cu. In this study it
was proved that the most appropriate chemical modier for all ele-
ments tested in all possible combinations was 1.25 lg Pd since it
provided signicant sensitivity and low background absorption.
The NH4H2PO4 and the permanent chemical modier ZrIr did
not provide signicant sensitivity for the determination of AsCu
and Cu, respectively. On the contrary, in the presence of PdMg
as chemical modier the background absorbance was increased
signicantly (Suppl. Fig. 2).
Using 1.25 lg Pd as a chemical modier, three different ap-
proaches were tested in order to minimise the analyses time and
cost: (a) the single element determination; (b) the by pairs ele-
ment determination (CdPb and CuAs); and (c) the simultaneous
determination of all four analytes. Table 1 shows the optimal com-
promised conditions and the calculated analytical characteristics
for all three different approaches. Among the approaches tested,
the by pairs element determination proved to be the most appro-
priate for the ofcial determination of Pb, Cd, As and Cu in seafood
and sh feed samples since it provides low detection limits and
1 neous determination of all four analytes did not seem to be an
appropriate approach for the ofcial control because it provides
high limits of detection due to the fact that the compromised con-
ditions were signicantly different than the respective conditions
in the single element approach and these limits were out of the
maximum permissible levels. For Pb and Cd in particular, the max-
imum permissible limits of detection in seafood samples according
Table 2
Comparison of the calibration curves obtained by aqueous standard solutions and
seafood matrix matched standard solutions for the simultaneous determination of
CdPb and AsCu. Pd was used as chemical modier.

Analyte Intercept SD Slope SD Correlation

There are many studies based on the use of permanent or non- coefcient (R2)
permanent modiers for the determination of different elemental Calibration with aqueous standard solutions
combinations in many food matrices (Ajtony et al., 2008; Caldas, Cd 0.0045 0.00032 0.0452 0.0014 0.998
Raposo, Gomes Neto, & Barbosa, 2009; Freschi, Freschi, & Neto, Pb 0.00050 0.00048 0.0025 0.00002 0.9998
Cu 0.0059 0.0014 0.003567 0.000093 0.998
2008). The objective of this study was to use a chemical modier
As 0.00149 0.00044 0.002053 0.000014 0.9998
capable of providing high sensitivity for the analytes tested. For
Calibration with matrix matched standard solutions
this reason, the non-permanent chemical modiers Pd, PdMg,
Cd 0.0038 0.00016 0.03730 0.00062 0.9992
NH4H2PO4, and the permanent chemical modier ZrIr were inves- Pb 0.00477 0.00032 0.00213 0.00001 0.9999
tigated as potential chemical modiers. The heating programme Cu 0.0149 0.0016 0.00362 0.000052 0.9994
for coating the integrated platform of a THGA with Zr followed As 0.02330 0.0034 0.001332 0.000011 0.9998
by Ir, was demonstrated by Lima, Krug, and Jackson (1998) and
 A.K. Psoma et al. / Food Chemistry 151 (2014) 7278
Table 3
Comparison of the calculated uncertainties and the maximum standard calculated
uncertainties obtained by three different concentration levels for the simultaneous
determination of CdPb and AsCu in seafood and sh feed samples.
Analyte concentration Calculated uncertainty Maximum uncertainty
level (ng/g) (ng/g) (ng/g)
Seafood samples
Cd (4.5280700) 0.863585 0.9552105
Pb (856706100) 9.568607 18102732
Cu (46010003000) 52108326 74125370
As (320130015000) 52143685 591531488
Fish feed samples
Cd (4.5280700) 0.823585 1.052105
Pb (856706100) 8.347417 23103417
Cu (46010003000) 83128325 89134373
As (320130015000) 561471280 601541489
to Commission Regulation 333/2007/EC are 30 and 5 lg kg1,
respectively, when the calculated limits were 40 and 3.2 lg kg1,
respectively. Furthermore, the limit of detection for the determina-
tion of As was increased at least 100 times compared to the single
element approach and for about 30 times compared to the by
pairs approach. On the other hand, the by pairs approach ele-
mental determination was considered the most appropriate, since
it provided similar sensitivity compared to the single element
determination and the limits of detection were less than the max-
imum permissible limits of detection in seafood samples according
to Commission Regulation 333/2007/EC. By using this approach,
the analysis time was reduced signicantly. For example, the anal-
ysis time for the determination of Pb, Cd, As and Cu in 100 samples
was reduced to about 8 h.
3.2. Method validation
Calibration curves for the simultaneous determination of CuAs
and CdPb were constructed using 1.25 lg of Pd for all the differ-
ent matrices. Curves in the 10.050.0 lg/L for As and Cu 5.00
50.0 lg/L for Pb, and 0.505.00 lg/L for Cd concentration range
were established using the integrated analyte absorbance versus
analyte concentration and typical linear correlations of
R2 P 0.998 were obtained. Quantication was also performed with
matrix-matched calibration curves, constructed from integrated
absorbance measurements and prepared automatically by the
autosampler by spiking 20 lL of a digested sample solution with
20 lL of different standard solutions (10.0, 20.0, 40.0 and
50.0 lg L1 for the simultaneous determination of Cu and As;
0.50, 1.00, 2.50 and 5.00 lg L1; and 5.00, 10.0, 25.0 and
50.0 lg L1 for the simultaneous determination of Cd and Pb). Re-
sults are summarised in Table 2. It was observed that, in the pres-
ence of the 1.25 lg of Pd, the slopes of the calibration curves and
the matrix matched calibration curves were different for Cd, Pb
and As (Table 2). This means that, for accurate results, calibration
with matrix-matched standards should be applied.
The instrumental LODs are presented in Table 1. Furthermore,
the methods LODs were 0.50, 18, 56 and 11 ng/g for the simulta-
neous determination of CdPb and CuAs in seafood samples,
respectively, and 1.0, 35, 112, and 22 ng/g for the simultaneous
determination of CdPb and CuAs in sh feed samples, respec-
tively. Finally, the calculated LOQs were 1.5, 53, 168, and 32 ng/g
for the simultaneous determination of CdPb and CuAs in seafood
samples and 3.0, 105, 336, and 65 ng/g for the simultaneous deter-
mination of CdPb and CuAs in sh feed samples, respectively.
The calculated LODs were lower than the suggested values as pre-
sented in Commission Regulation 333/2007/EC and the maximum
permissible levels as presented in Regulations (EC) No. 1881/2006
75
(<30, and <5 lg kg1 for the determination of Pb and Cd in seafood
samples, respectively) and (EC) No. 32/2002 (<500, <50, and
<400 lg kg1 for the determination of Pb, Cd and As in sh feed
samples, respectively).
Precision experiments were carried out and the relative stan-
dard deviation (%RSD) values achieved from three different con-
centration levels, measured six times under repeatability
conditions and six times at two different days under reproducibil-
ity conditions, were lower than 10% for all different concentration
levels and for all analytes. The HORRATr and HORRATR values
achieved from these different concentration levels, ranged from
0.27 to 0.48 for Cd, 0.19 to 0.50 for Pb, 0.51 to 0.67 for As and
0.30 to 0.54 for Cu. These values were lower than the crucial value
of two, as described by the European Commission Regulation 333/
2007 for the ofcial control of the levels of lead, cadmium and
other contaminants in foodstuffs. However, this regulation may
be used also when a limited number of fully validated methods
of analysis exist, and a tness-for-purpose approach may be used
to assess the suitability of the method of analysis.
For accuracy estimation additional recovery experiments were
carried out in order to check the trueness of the proposed method.
The (%) calculated recoveries for Cd, Pb, As and Cu, expressed as the
mean value of six different spiked samples for each different con-
centration level ranged from 86% to 104% for sh feed samples and
95108% for seafood samples. The recovery data are within 20% of
the target value (European Commission, 2008a) and for this reason
the developed method was considered as t for purpose.
The uncertainty of the method was also calculated based on the
Eurachem/Citac Guidelines. In practice, the uncertainty of the re-
sults in this study arose from many possible sources, including ma-
trix effects and interferences, environmental conditions,
uncertainties of masses and volumetric equipment, reference val-
ues, approximations and assumptions incorporated in the mea-
surement method and procedure, and random variation. The
combined uncertainty uc(y) was calculated from the summary
squared of several independent parameters such as: (a) the mass
uncertainty; (b) the mix stock standard solution; (c) the volume
uncertainty; (d) the calibration uncertainty; (e) the bias uncer-
tainty, as estimated by the addition recovery tests and through
the comparison of the calculated recoveries and the theoretical
ones provided by the Commission Regulation 333/2007/EC
(100 20%), and (f) the precision uncertainty, as estimated by the
% RSDR values for the three different concentration levels under
reproducibility conditions. The choice of the factor k is based on
the level of condence desired. For an approximate level of con-
dence of 95%, k is 2. The calculated expanded uncertainties and
the respective theoretical one calculated by Eq. (1) are shown in
Table 3. The results show that the method is suitable for the ofcial
control of Cd, Pb, As and Cu in sh and sh feed samples due to the
fact that it produced results with standard measurement uncer-
tainties less than the maximum standard measurement
uncertainty.
3.3. Determination of Cd, Pb, As and Cu in seafood and sh feed
samples
The developed and accredited method was applied for the
simultaneous determination of Cd and Pb in 19 different seafood
species (totally 202 samples) and the results were compared to
the European Regulation (EC) No. 1881/2006. Furthermore, the
method was also applied for the determination of Cd, Pb, As and
Cu in 275 sh feed samples and the results were compared to
the Directive 2002/32/EC. Finally, an overall seasonal variation of
metals and metalloid accumulation for over 8 years in the two
most produced shes in Greece (sea bream and sea bass) was
accomplished for the rst time.
76 A.K. Psoma et al. / Food Chemistry 151 (2014) 7278

Table 4
Cd and Pb content in seventeen different seafood species purchased from the Greek market.

Seafood species No. Cd (lg/kg) Pb (lg/kg)

samples
xmean SD xminxmax (EC) No. 1881/ xmean SD xmin (EC) No. 1881/
2006 xmax 2006
Shrimp (Parapenaeus longirostris) 29 23 30 <0.5165 500 38 22 <1889 500
Red snapper (Lutjanus campechanus) 4 1.7 1.6 <0.53.2 50 13.4 8.9 <18 300
26.7
Atlantic cod (Gadus morhua) 9 2.2 3.2 <0.59.8 50 24 16 <1851 300
Surmullet (Mullus surmuletus) 4 14 11 <0.528 50 19 20 <1849 300
Common pandora (Pagellus erythrinus) 2 4.4 2.26.5 50 29.2 26.1 300
32.4
Swordsh (Xiphias gladius) 5 5.4 5.3 <0.514.3 300 23 13 <1836 300
Mussels (Mytilus edulis) 4 4.2  102 2.7  102 1.6  102 1000 47 40 <1888 1500
7.2  102
Blue shark (Prionace glauca) 4 1.2 1.9 <0.54.0 50 19 17 <1838 300
Red porgya (Pargus pargus) 5 0.9 1.5 <0.53.7 50 <18 <18<18 300
Percha (Perca uviatilis) 4 <0.5 <0.5<0.5 50 24 17 <1838 300
European pilcharda (Sardina pilchardus) 1 3.8 100 53 300
Solea (Solea Solea) 2 <0.5 <0.5<0.5 50 46 <1883 300
Cuttlesha (Sepia latimanus) 2 28 39 <0.556 1000 <18 <18<18 1000
Squida (Loligo vulgaris) 2 3.6  102 4.0  102 0.8  102 1000 <18 <18<18 1000
6.4  102
Octopus (Octopus vulgaris) 3 36 54 <0.598 1000 52 44 <1897 1000
Longnose spurdoga (Squalus blainville) 1 4.07 50 47.6 300
Sharpsnout sea breama (Diplodus 1 3.17 50 63.78 300
puntazzo)
a
For some samples only the mean value is provided either due to the small number of the samples (one or two) or due to the fact that the minimum and maximum content
of the analytes were lower that the limit of detection.

The samples were obtained from local markets in Athens,
Greece and different aquaculture centres (from 2005 to 2013).
Samples were milled and homogenised in a food processor with
a stainless steel cutter to obtain a representative material. Proce-
dural blanks were always run and contamination from this step
was never observed. Each sample was digested and measured
three times. Contents were calculated on a wet weight basis. As
shown in Tables 4 and 5, the contents of Cd, Pb in the muscle tissue
of the different sh samples (n = 162, from 14 different seafood
species) ranged from <0.5 lg/kg to 28 lg/kg and <18 lg/kg to
162.6 lg/kg, respectively. As concerns the content levels of Cd
and Pb in crustaceans (shrimps), they ranged from <0.5 lg/kg to
165 lg/kg, n = 29 and <18 lg/kg to 89 lg/kg, n = 29, respectively.
Furthermore, bivalve molluscs (mussels) showed the highest con-
tent of Cd and Pb among all other seafood species, as expected,
and the values ranged from 0.16 mg/kg to 0.72 mg/kg, n = 4 and
<18 lg/kg to 88 lg/kg, n = 4, respectively. Finally, in cephalopods
(squids, cuttleshes, octopus) content of Cd and Pb ranged from
<0.5 lg/kg to 640 lg/kg, n = 7 and <18 lg/kg to 97 lg/kg, n = 7,
respectively. These values were lower than the crucial values, as
described by the European Commission Regulation 1881/2006 set-
ting maximum levels for certain contaminants in foodstuffs. Com-
paring the quality of the consumed seafood species in Greece to
those in other countries as reported in the literature, it can be pro-
posed that Greek seafood samples tend to have a lower content of
toxic elements. Generally, cadmium and lead levels in analysed sh
samples, as described in the literature, were found to be similar or
higher than the maximum permissible limits. For example, Tuzen
(2009) found 0.280.87 mg/kg for lead and 0.100.35 mg/kg for
cadmium in ten different seafood species from the Black Sea. Sim-
ilar results were reported from Trkmen et al. (2010) who found
0.100.47 mg/kg cadmium, 0.190.47 mg/kg for lead in sea bream
(Sparus aurata), European sea bass (Dicentrarchus labrax), and
keeled mullet (Liza carinata Valenciennes) from the Mediterranean
region. High metal concentrations in edible parts of seafood species
were also reported from Trkmen et al. (2008) (0.020.37 mg/kg
for cadmium and 0.330.86 mg/kg for lead, respectively) from
Marmara, Aegean and Mediterranean seas in Turkey. On the
contrary, the concentrations found for cadmium and lead from
Prez Cida, Boia, Pombo, and Rebelo (2001), in the muscle of the
Anguilla anguilla, Mullus surmuletus, Trigla lucerna, Mugil cephalus,
Chelon labrosus, Liza aurata and D. labrax from Portugal were lower
than 0.043, 0.15 respectively.
On the contrary, As and Cu, which are not included in the Euro-
pean Commission Regulation 1881/2006, were determined only for
the most consumed seafood samples in Greece (sea bream and sea
bass). These sh species showed that As content ranged from
0.038 mg/kg to 3.85 mg/kg, n = 59 and Cu content ranged from
0.17 mg/kg to 1.74 mg/kg, n = 78, respectively (Table 5). Several
studies have proven that the greatest part of arsenic is in the or-
ganic form (over 68%) (Julshamn et al., 2012; Schaeffer, Soeroes,
Ipolyi, Fodor, & Thomaidis, 2005; Shah et al., 2010). Taking into ac-
count that the greatest portion of As due to dietary exposure (94%
of the total As) comes from the seafood group (Ysart et al., 2000)
and that a provisional tolerable weekly intake (PTWI) for total As
has not been recommended, a further study for the potential risks
for human health should be undertaken. However, in a previous
study for As speciation in Aegean Sea seafood, no risks associated
to As dietary exposure were found (Schaeffer et al., 2005). Never-
theless, to assess the risk for consumers from copper intake
through sh consumption, the weekly intake for a 70-kg adult
was compared with the PTWI (3.5 mg Cu/kg body weight/week)
(FAO/WHO, 2004). According to the data available through the
European programme DAFNE (Data Food Network) based on the
average diet per person, the daily average sh consumption in
Greece is equal to 46 g. This means that to overcome the PTWI va-
lue, a 70-kg adult must consume 570 kg of seafood/week. Compar-
ing the results obtained from the current study to others presented
in the literature, it can be assumed that the results are in agree-
ment with values reported by many researchers for seafood sam-
ples (Prez Cida et al., 2001; Trkmen et al., 2008, 2010; Tuzen,
2009).
Similarly, the contents of Cd, Pb and As in sh feed samples
were lower than the maximum permissible values, as described
by Directive 2002/32/EC concerning the undesirable substances
in animal feeds. More specically, the content of Cd ranged from
 A.K. Psoma et al. / Food Chemistry 151 (2014) 7278 77

Table 5
Cd, Pb, As and Cu content in sea bream and sea bass sh samples obtained from Greek aquacultures.

Sea bream (Sparus aurata) Sea bass (Dicenthrarchus labrax)

Cd (lg/kg) No. samples 53 67
xmean 2.3 1.9
SD 3.1 2.2
xminxmax <0.518.6 <0.59.5
Pb (lg/kg) No. samples 53 67
xmean 43 38
SD 39 38
xminxmax <18137.5 <18162.6
As (mg/kg) No. samples 29 30
xmean 0.95 0.51
SD 0.73 0.33
xminxmax 0.043.85 0.061.23
Cu (mg/kg) No. samples 39 39
xmean 0.46 0.43
SD 0.29 0.26
xminxmax 0.171.74 0.191.53

concentration can be exported. A Spearmans non-parametric cor-

Table 6 relation was performed, using IBM SPSS Statistics 20 statistical
Cd, Pb, As and Cu content in different sh feed samples obtained from Greek tool, in order to interpret the years and the months with the ele-
aquacultures from 2005 to 2013.
mental content. The Spearman rank correlation method makes
Fish feed samples no assumptions about the distribution of the data. It may therefore
Cd (mg/kg) No. samples 200 be more appropriate for data with large outliers that hide mean-
xmean 0.29 ingful relationships between series or for series that are not nor-
SD 0.11 mally distributed. The results show the potential of the
xminxmax 0.0010.490
chemometric technique to track the correlation of seasonal varia-
Pb (mg/kg) No. samples 275 tion to the sh samples elemental content. Taking into account
xmean 0.35
SD 0.49
the calculated coefcient, it can be assumed that there is no corre-
xminxmax <0.0350.43 lation among the months and the elemental content for the ana-
As (mg/kg) No. samples 247
lysed sh samples for all these 8 years (Suppl. Table 1). This
xmean 1.27 proves that the different seasonal conditions did not affect the ele-
SD 0.93 mental content of the sh samples. This conclusion comes on con-
xminxmax <0.023.49 trary with the ndings of Ekpo and Ibok (1999), Obasohan and
Cu (mg/kg) No. samples 259 Eguavoen (2008) and Mendil et al. (2010) who estimated that the
xmean 7.9 differences in the seasonal levels for most of the metals were sig-
SD 4.2
nicant. However, these studies provide conclusions only about
xminxmax 0.223.1
the correlation between the seasonal variability and the elemental
accumulation for a 1 year period.
In addition, a signicant correlation among the sampling year
0.001 mg/kg to 0.49 mg/kg, n = 200, the content of Pb ranged from and the content of Pb (0.352, n = 64, CL = 99%), Cu (0.494,
<0.035 mg/kg to 0.43 mg/kg, n = 275, and As content ranged from n = 26, CL = 95%) and As (0.474, n = 26, CL = 95%) in sea bass
<0.02 mg/kg to 3.49 mg/kg, n = 247, respectively. On the other and the content of Cd (0.342, n = 53, CL = 95%) and As (0.562,
hand, the Cu content was equal to (7.9 4.2) mg/kg, n = 259 (Ta- n = 26, CL = 99%) in sea bream was proved (Suppl. Table 1). This
ble 6). Guidelines from the Association of Feed Control Ofcials correlation is directly depended from the amount of the elemental
(Hanks, 2000) indicate the maximum tolerable level for Cu equal intake through different sh feeds used in aquacultures, as proved
to 25 mg/kg. Maule et al. (2007), have determined several metal by this study concerning the fact that the content of Pb, Cd, As in
contaminants in 55 sh feed samples assayed by Severn Trent Lab- sh feeds was also negatively correlated with the years (0.425,
oratories, Inc. and the National Water Quality Laboratory in sh n = 200, CL = 99%; 0.154, n = 250, CL = 95%; 0.423, n = 247,
feed samples collected from 11 National Fish Hatcheries between CL = 99%, respectively) and Cu was also positively correlated
October 2001October 2003 and the results were in accordance (0.308, n = 259, CL = 99%) (Suppl. Table 1).
with those presented in the current study (As = 2.62 mg/kg;
Cd = 0.39 mg/kg; Pb = 0.78 mg/kg; Cu = 10.54 mg/kg) (Maule
et al., 2007). As Maule et al. (2007) also concluded, these dietary 4. Conclusion
levels in the feed had not affected the sh quality, as can be seen
by the low content of toxic metals in sh samples. The current work described a method for the simultaneous
determination of PbCd and AsCu in seafood (n = 202) and sh
3.4. Seasonal correlation of the investigated trace elements feed (n = 275) samples by Simultaneous Electrothermal Atomic
Absorption Spectrometry (SIMAAS) in regard to Council Directive
In this study an overall seasonal variation of metal accumula- 333/2007 EC and ISO/IEC 17025 (2005). The method was accred-
tion for over 8 years and for sea bream and sea bass sh samples ited and implemented in routine analysis in order to investigate
was accomplished for the rst time. Due to the great number of the seafood quality in Greek market for the last 8 years. The results
collected data sets, reliable conclusions about the correlation indicated that the elemental contents were lower than the legisla-
between the temporal (monthly) variability and the element tive values, as described by the European Commission Regulation
78 A.K. Psoma et al. / Food Chemistry 151 (2014) 7278
1881/2006 setting maximum levels for certain contaminants in
foodstuffs. Finally, it was proved that there is no seasonal variation
of metal content in sea bream and sea bass sh samples and that
there is a strong correlation among the sh dietary and the ele-
mental accumulation.
Appendix A. Supplementary data
Supplementary data associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.foodchem.
2013.11.045.
References
Ajtony, Z., Szoboszlai, N., Susk, E. K., Mezei, P., Gyrgy, K., & Bencs, L. (2008). Direct
sample introduction of wines in graphite furnace atomic absorption
spectrometry for the simultaneous determination of arsenic, cadmium,
copper and lead content. Talanta, 76, 627634.
Caldas, N. M., Raposo, J. L., Jr., Gomes Neto, J. A., & Barbosa, F. Jr., (2009). Effect of
modiers for As, Cu and Pb determinations in sugar-cane spirits by GFAAS. Food
Chemistry, 113, 12661271.
DAFNEANEMOS. (2013). Data food networking. <http://www.hhf-greece.gr/
dafnesoftweb/Main.aspx?type=multi> Accessed 06.06.13.
Ekpo, B. O., & Ibok, U. J. (1999). Temporal variation and distribution of trace metals
in freshwater and sh from Calabar River, S.E. Nigeria. Environmental
Geochemistry and Health, 21, 5166.
Eurachem Citac Guide CG4. (2000). Quantifying Uncertainty in Analytical
Measurement (2nd ed.), English edition.
European Commission. (2006). Regulation (EC) No. 1881/2006 of 19 December 2006
setting maximum levels for certain contaminants in foodstuffs, L 364/5L 364/24.
European Commission. (2007). Regulation (EC) No. 333/2007 of 28 March 2007 laying
down the methods of sampling and analysis for the ofcial control of the levels of
lead, cadmium, mercury, inorganic tin, 3-MCPD and benzo(a)pyrene in foodstuffs, L
88/29L 88/38.
European Commission. (2008a). Regulation (EC) No. 629/2008 of 2 July 2008
amending Regulation (EC) No. 1881/2006 setting maximum levels for certain
contaminants in foodstuffs, L 173/6L 173/9.
Regulation (EC) No. 765/2008 of the European Parliament and of the Council of 9
July 2008b setting out the requirements for accreditation and market
surveillance relating to the marketing of products and repealing Regulation
(EEC) No. 339/93, L 218/30L 218/47.
European Commission. (2002). Decision of 12 August 2002 implementing Council
Directive 96/23/EC concerning the performance of analytical methods and the
interpretation of results, L 221/8L 221/36.
European Commission. (2002). Directive 2002/32/EC of the European Parliament and
of the Council of 7 May 2002 on undesirable substances in animal feed, L 140/10
L140-21.
FAO/WHO. (2004). Summary of evaluations performed by the Joint FAO/WHO Expert
Committee on Food Additives (JECFA 19562003), (First through sixty-rst
meetings). ILSI Press International Life Sciences Institute.
Freschi, G. P. G., Freschi, C. D., & Neto, J. A. G. (2008). Evaluation of different rhodium
modiers and coatings on the simultaneous determination of As, Bi, Pb, Sb, Se
and of Co, Cr, Cu, Fe, Mn in milk by electrothermal atomic absorption
spectrometry. Microchimica Acta, 161, 129135.
Hanks, A. (2000). Mineral feed contaminants pp. 265269. Oxford, IN: Ofcial
Publication of the Association of Feed Control Ofcials Inc..
ISO/IEC 17025. (2005). General requirements for the competence of testing and
calibration laboratories.
Julshamn, K., Nilsen, B. M., Frantzen, S., Valdersnes, S., Maage, A., Nedreaas, K., et al.
(2012). Total and inorganic arsenic in sh samples from Norwegian waters.
Food Additives and Contaminants: Part B: Surveillance, 5, 229235.
Kalantzis, V., Rousis, N., Pasias, I., Thomaidis, N., & Piperaki, E. (2012). Evaluation of
different modiers for the determination of arsenic in leachate samples from
sanitary landlls by electrothermal atomic absorption spectrometry. Analytical
Letters, 45, 592602.
Lima, E. C., Krug, F. J., & Jackson, K. W. (1998). Evaluation of tungstenrhodium
coating on an integrated platform as a permanent chemical modier for
cadmium, lead, and selenium determination by electrothermal atomic
absorption spectrometry. Spectrochimica Acta Part B, 53, 17911804.
Low, K. H., Zain, S. Md., & Radzi Abas, Mhd. (2012). Evaluation of microwave-
assisted digestion condition for the determination of metals in sh samples by
inductively coupled plasma mass spectrometry using experimental designs.
International Journal of Environmental Analytical Chemistry, 99, 11611175.
Maule, A. G., Gannam, A. L., & Davis, J. W. (2007). Chemical contaminants in sh
feeds used in federal salmonid hatcheries in the USA. Chemosphere, 67,
13081315.
Mendil, D., Demirci, Z., Tuzen, M., & Soylak, M. (2010). Seasonal investigation of
trace element contents in commercially valuable sh species from the Black sea,
Turkey. Food and Chemical Toxicology, 48, 865870.
Obasohan, E. E., & Eguavoen, O. I. (2008). Seasonal variations of bioaccumulation of
heavy metals in a freshwater sh (Erpetoichthys Calabaricus) from Ogba River,
Benin city, Nigeria. Indian Journal of Animal Research, 42, 171179.
Prez Cida, B., Boia, C., Pombo, L., & Rebelo, E. (2001). Determination of trace metals
in sh species of the Ria de Aveiro, (Portugal) by electrothermal atomic
absorption spectrometry. Food Chemistry, 75, 93100.
Schaeffer, R., Soeroes, C., Ipolyi, I., Fodor, P., & Thomaidis, N. S. (2005).
Determination of arsenic species in seafood samples from the Aegean Sea by
liquid chromatography (Photo-oxidation) Hydride generation Atomic
uorescence spectrometry. Analytica Chimica Acta, 547, 109118.
Shah, A. Q., Kazi, T. G., Baig, J. A., Afridi, M. B. A. H. I., Kandhro, G. A., Wadhwa, S. K.,
et al. (2010). Determination of inorganic arsenic species (As3+ and As5+) in
muscle tissues o sh species by electrothermal atomic absorption spectrometry
(ETAAS). Food Chemistry, 119, 840844.
Taverniers, I., De Loose, M., & Van Bockstaele, E. (2004). Trends in quality in the
analytical laboratory. II. Analytical method validation and quality assurance.
Trends in Analytical Chemistry, 23, 535552.
Thompson, M. (2000). Recent trends in inter-laboratory precision at ppb and sub-
ppb concentrations in relation to tness for purpose criteria in prociency
testing. Analyst, 125, 385386.
Trkmen, M., Trkmen, A., Tepe, Y., Ates, A., & Gkkus, K. (2008). Determination of
metal contaminations in sea foods from Marmara, Aegean and Mediterranean
seas: Twelve sh species. Food Chemistry, 108, 794800.
Trkmen, A., Trkmen, M., Tepe, Y., & eki, M. (2010). Metals in tissues of sh from
Yelkoma Lagoon. Environmental Monitoring Assessment, 168, 223230.
Tuzen, M. (2009). Toxic and essential trace elemental contents in sh species from
the Black Sea. Turkey Food and Chemical Toxicology, 47, 17851790.
Ysart, G., Miller, P., Croasdale, M., Crews, H., Robb, P., Baxter, M., et al. (2000). UK
total diet study-dietary exposures to aluminum, arsenic, cadmium, chromium,
copper, lead, mercury, nickel, selenium, tin and zinc. Food Additives and
Contaminants, 17, 775786.