Anal Bioanal Chem (2011) 400:2185–2193

DOI 10.1007/s00216-011-4945-z

ORIGINAL PAPER

Multiresidue method to quantify pesticides in fish muscle

by QuEChERS-based extraction and LC-MS/MS
Angélique Lazartigues & Laure Wiest & Robert Baudot &
Marielle Thomas & Cyril Feidt & Cécile Cren-Olivé

Received: 6 January 2011 / Revised: 25 February 2011 / Accepted: 22 March 2011 / Published online: 12 April 2011
# Springer-Verlag 2011

Abstract Pesticide residues in fish muscle are an Introduction

environmental and a health safety concern which requires
analytical methods presenting high sensitivity and low Agricultural activities induced a pesticide contamination of
limits of quantification. In this study, adapted QuEChERS numerous aquatic ecosystems [1–3]. Nowadays, research
method, coupled to liquid chromatography tandem mass on contamination of surface waters or sediments has
spectrometry (Scheduled MRM-5500 QTRAP), was become a major preoccupation in environmental science.
developed to quantify 13 pesticides (azoxystrobin, clomazone, A consequence of this widespread contamination, in which
diflufenican, dimethachlor, carbendazim, iprodion, isopro- pesticides were often combined in a mixture, was the
turon, mesosulfuron-methyl, metazachlor, napropamid, impact on fish health [4–6] or accumulation in edible tissue
quizalofop and thifensulfuron-methyl) in muscle of fish. of fish [7], with potentially a risk for consumers [8, 9].
Quantification limits were below 1 ng g−1 except for Although a large range of pesticides could be potentially
clomazone (1.8 ng g−1) and quizalofop (7.4 ng g−1). Best transferred into fish muscle, few of them are investigated.
recoveries were observed for perch (>80%) and roach In light of these concerns, evaluation of levels of pesticides
(>68%), except for thifensulfuron-methyl. Lower recoveries in muscle of fish is an important objective for environmental
had been observed for carp (6% to 86%). Relative standard and health sciences.
deviation was lower than 28% for intra-day and 29% for Multiresidue approaches, with low quantification limits,
inter-day analysis, respectively. This method was successfully are a rapid methodological answer. Chromatography cou-
tested on three fish species, naturally or orally exposed: roach pled with mass spectrometry (MS) is generally used for
(Rutilus rutilus), perch (Perca fluviatilis) and carp (Cyprinus determination of pesticides in food or in environmental
carpio). Few levels were observed in fish naturally exposed, matrices [10–16]. Liquid chromatography (LC) coupled to
but carp and perch orally contaminated showed measurable MS is generally used to explore a large range of pesticides
levels in their muscles. [17], from low volatile to medium-high polar compounds. It
is the method of choice regarding water and food samples
Keywords Multiresidue . QuEChERS . Pesticides . Fish . [10, 18–21].
LC-MS/MS For complex matrices, an extraction phase development
is generally required. A lot of methods were developed to
A. Lazartigues : M. Thomas : C. Feidt extract compounds from fish muscle. Solid–liquid extrac-
Unité de Recherche Animal et Fonctionnalité des Produits tion was sometimes used but concerned generally one
Animaux, Nancy Université, INRA,
2 avenue de la Forêt de Haye, BP 172,
family of compound with similar properties. Pressurized
54505 Vandœuvre-lès-Nancy, France liquid extraction was also applied [22], but this method was
most developed in the case of pharmaceuticals. Microwave-
L. Wiest : R. Baudot : C. Cren-Olivé (*) assisted extraction was a generic method to extract
Service Central d’Analyse du CNRS – USR59,
Echangeur de Solaize Chemin du Canal,
compounds in animal tissues, but reproducibility was
69360 Solaize, France moderate depending on dispersion of microwaves. Quick,
e-mail: [email protected] easy, cheap, effective, rugged and safe method (QuEChERS)
2186 A. Lazartigues et al.

[23] seemed to be a right way to extract pesticides from contaminated in environmental conditions, carp (Cyprinus
complex matrices, providing high quality results with a carpio) and perch (Perca fluviatilis), orally contaminated in
minimum number of steps. It was especially true for fruits laboratory. These three species were explored due to their
and vegetables [24] or wheat grain and flour [25]. For some economic importance in freshwater ponds production and
matrices, this method needed modifications as it was recently also their importance to local human fish consumption.
demonstrated for tea [26] or for banana leaves [27]. A recent
QuEChERS method was also developed for pyrethrin and
pyrethroid pesticides in fish [28]. However, complementary Materials and methods
researches are necessary for the determination of a large
range of pesticides in higher fat-containing matrices as fish Chemicals and reagents
[24], which could contain polar and non-polar residues [29].
In this context, the aim of this work was the develop- Acetonitrile and hexane were of high purity (99%) and
ment of a multiresidue method for the simultaneous purchased from Sigma-Aldrich (Saint Quentin Fallavier,
analysis of a list of 13 pesticides in three freshwater fish. France). Dimethyl sulfoxyde (DMSO) was of high purity
These compounds demonstrated a large range of physico- (99%) and obtained from Riedel-de haen (Seelze, Germany).
chemical properties (Table 1). They were also among the Ultrapure (UP) water was produced from a Direct Q-system
most commonly applied pesticides in France for cereals or (Millipore, Molsheim, France). Azoxystrobin, clomazone,
oleaginous crop [30] (i.e. azoxystrobin, clomazone, diflu- diflufenican, dimethachlor, carbendazim, iprodion, isopro-
fenican, dimethachlor, carbendazim, iprodion, isoproturon, turon, mesosulfuron-methyl, metazachlor, napropamid,
mesosulfuron-methyl, metazachlor, napropamid, quizalo- thifensulfuron-methyl (Sigma-Aldrich) and quizalofop
fop, thifensulfuron-methyl). Method was based on ultra- (Supelco, France) were from 97% to 99% purity. Stock
turrax sample homogenization and extraction adapted from solutions of each compound at 1 mg/mL except for
QuEChERS method followed by a LC-MS/MS analysis. carbendazim (0.1 mg/mL, in methanol, because of limited
Matrix effects were investigated and compared between the solubility) were prepared in acetonitrile and stored at −20 °C.
three different fish species. This method exhibits signifi- Diluted solutions from 1 ng L−1 to 50 μg L−1 were used for
cantly improved sensitivity due to the QqLIT (hybrid triple calibration curves.
quadrupole linear ion trap) capabilities and very low QuEChERS materials (Agilent SampliQ QuEChERS
quantification limits compared to the previously published Kits) were purchased from Agilent Technologies. The
methods for determination of pesticides in animal muscle citrate-buffered version was chosen: a mixture contain-
[31]. This method was applied on muscle of three ing 4 g anhydride magnesium sulphate (MgSO4), 1 g
freshwater fish: roach (Rutilus rutilus), which had been sodium chloride (NaCl), 1 g trisodium citrate dihydrate

Table 1 Properties of the 13 pesticides

Compounds Formulae Actiona Cropb Molar massc Solubility in Vapour pressurec Log Kowc
(g/mol) waterc (mg/L) (Pa, 20 °C)

Azoxystrobin C22H17N3O5 F Wheat/barley 403.39 6.7 1.1×10−10 2.50

Carbendazim C9H9N3O2 F Rape seed 191.19 8 9.0×10−5 1.56
Clomazone C12H14ClNO2 H Rape seed 239.70 1,100 1.9×10−2 2.54
Diflufenican C19H11F5N2O2 H Wheat/barley 394.29 0.05 3.1×10−5 4.90
Dimethachlor C13H18ClNO2 H Rape seed 255.74 2,300 1.5×10−3 2.25
Fluroxypyr C7H5Cl2FNO3 H Wheat/barley 255.03 6,500 3.8×10−9 2.00
Iprodion C13H13Cl2N3O3 F Rape seed 330.17 13 5.0×10−7 2.99
Isoproturon C12H18N2O H Wheat/barley 206.28 70 3.3×10−6 2.50
Mesosulfuron-methyl C17H21S2N5O9 H Wheat 503.51 483 1.1×10−11 −0.48
Metazachlor C14H16ClN3O H Rape seed 277.75 450 4.7×10−5 2.13
Napropamid C17H21NO2 H Rape seed 271.35 70 5.3×10−4 3.36
Quizalofop C17H13ClN2O4 H Rape seed 344.80 0.31 1.1×10−7 4.66
Thifensulfuron-methyl C12H13S2N5O6 H Maize 387.39 2,240 7.5×10−9 −1.7
a
Action is symbolized by H for herbicide and F for fungicide action
b
Examples of crop are presented to illustrate pesticide use (Couteux [30])
c
Data come from pesticide databases (data confirmed by four web databases: SIRIS, Footprint, Agritox and HSDB)
Multiresidue method to quantify pesticides in fish muscle 2187

(Na 3 Cit·2H 2 O) and 0.5 g disodium hydrogencitrate acetonitrile/water (10:90, v/v). Finally, the sample was
sesquihydrate (Na2HCit·1.5H2O) was employed. capped, dark stored (−20 °C) and vortexed thoroughly
before LC-MS/MS analysis
Sample collection
LC-MS/MS analysis
Depuration in clean water was performed for roaches (five
dam ponds in Lorraine region, France; 3 months depuration), Analyses were performed on Agilent 1290 series (Agilent
carps (“GAEC du Saulnois”, France; 3 months depuration) Technologies, Heilbronn, Germany) equipped with a
and perches (“UR-AFPA”, France; from egg to experimental reversed-phase Zorbax Eclipse C18 (Agilent, 50×2.1 mm,
juvenile phase life in clean water) for method development. 1.8 μm particle size). Mobile phases were 0.0125% (v/v, pH=
Other roaches were collected in five dam ponds in the East of 4.04) acetic acid (A) and 100% acetonitrile (B). Pesticides
France (Lorraine region) between 2007 and 2009 to represent were separated following gradient program: initial conditions
environmental samples. Some carps and perches had been were 10% B, then gradient was from 10% B to 100% B in
orally exposed via contaminated feed - to a mixture of the 13 5 min and finally, solvents were maintained to 100% B for
pesticides studied - in laboratory during 1 to 6 days. It 5 min. The equilibration time was 7 min, which resulted in a
represented 52 fishes per specie at different times during total run time of 17 min. Column temperature was 60 °C;
exposure (fish sampled at 0, 1, 2, 4, 5 and 6 days of exposure) flow rate was 0.3 mL min−1. The sample injection volume
or depuration (fish sampled at 0.375, 1, 1.375, 3, 6 and was set at 10 μL.
12 days of depuration) phases. For each time of sample, four The mass spectrometer used was a 5500 QTRAP system
repetitions - corresponding to four different fish exposed in the (Applied Biosystems, Foster City, CA, USA) equipped with
same conditions - were analyzed. an ESI (TurboV) source operated in positive ionization
Fishes were killed by a blow to the head, following mode. The operating parameters were ionspray voltage
anaesthesia in ice bath. Then, muscle were quickly removed 5,500 V, curtain gas 20 (arbitrary units), nebulizer gas 50
and frozen (−20 °C). All fish treatments used to obtain fish (arbitrary units), auxiliary gas 60 (arbitrary units) and probe
collection were in accordance with the general guidelines of temperature 600 °C. For each compound, declustering
the Council of European Communities [32] and the French potential (DP) and collision energy (CE) of the main
Animal Care Guidelines. transitions were optimised from a continuous flow of a
standard injection (100 μg L−1 solution at 10 μL min−1) to
Extraction method obtain the maximum intensities. Two MS/MS transitions
were acquired for each analyte, using the intensity ratio
A QuEChERS extraction procedure was developed for as confirmatory parameter. Scheduled MRM was opti-
the fish muscle having a weight of 3 g. Fish samples mized and used as follows: MRM detection window 60 s
were thawed out at +4 °C, grinded, and 3 g of muscle and target scan time 0.5 s. Each extract was injected
was sampled in 50 mL polypropylene centrifuge tube. twice in order to check the injection repeatability on the
Four millilitres of acetonitrile was added. A high- same column. LC-MS/MS conditions are presented in
performance disperser, ultra-turrax (IKA, Lille, France), Table 2.
was used to grind the sample in tubes maintained in
water at 20 °C. Ultra-turrax was then rinsed with 6 mL Validation
of acetonitrile.
One millilitre of ultrapure water, 3 mL of hexane and To determine the recoveries, fish samples were spiked in
200 μL of internal standards solution at 50 μg L−1 triplicate with a standard mixture of analytes at 50 μg L−1. The
(diflufenican-d3, carbendazim-d4 and isoproturon-d3) were spiked samples were stirred vigorously in order to enable
added, and the mixture was shaken during 30 s. Next, better contact of analytes with the matrix. These three
citrate salts (Agilent SampliQ QuEChERS extraction) was samples together with the correspondent blank samples were
added directly in the tube; the mixture was immediately and extracted and treated by the previously described protocol.
manually shaken to avoid the agglomeration of salts and The recoveries were determined in triplicate comparing the
was vigorously shaken during 1 min. obtained concentrations between the spiked-extracted sample
A centrifugation (5,000 rpm, 2 min, 20 °C) was and the blank fish samples spiked at the same concentration
performed, and upper phase (hexane) was removed. Six after the QuEChERS extraction.
millilitres of the second liquid phase (acetonitrile) was Limits of detection and quantification (LOD and LOQ)
transferred into a glass tube, and 50 μL of DMSO was were determined as the analyte concentration that produced
added. After evaporation (under nitrogen stream at 40 °C) a peak signal of three and ten times the background noise from
until 50 μL is left, residue was reconstituted in 2,950 μL of the chromatogram, respectively. The quantification was based
2188 A. Lazartigues et al.

Table 2 LC-MS/MS (Scheduled MRM-5500 QTRAP) conditions applied in positive ES mode: retention time (RT), transitions selected,
declustering potential (DP), collision energy (CE) and limits of detection (LOD) and quantification (LOQ) associated for roach, perch and carp

Compounds RT Transitions DP CE Roach Perch Carp

(min) (V) (V)
LOD LOQ LOD LOQ LOD LOQ
(ng g−1) (ng g−1) (ng g−1) (ng g−1) (ng g−1) (ng g−1)

Azoxystrobin 3.7 404.2→372.1 60 19 0.004 0.013 0.004 0.012 0.001 0.004

3.7 404.2→344.2 60 27
Carbendazim 1.6 192.2→160.2 71 31 0.005 0.034 0.004 0.009 0.005 0.009
1.6 192.2→132.2 71 43
Clomazone 3.4 240.1→125.1 60 23 0.028 0.093 0.018 0.059 0.020 0.067
3.4 240.1→89.0 60 59
Diflufenican 4.5 395.1→266.1 60 39 0.215 0.539 0.117 0.313 0.100 0.267
4.5 395.1→246.1 60 41
Dimethachlor 3.4 256.2→224.1 60 17 0.003 0.006 0.002 0.004 0.002 0.005
3.4 256.2→148.2 60 33
Fluroxypyr 2.3 255.0→209.0 66 21 0.956 2.079 0.421 0.915 0.384 1.253
2.3 255.0→181.0 66 33
Iprodion 3.9 330.1→245.2 46 21 0.681 1.346 0.914 1.805 1.219 2.407
3.9 330.1→288.1 46 17
Isoproturon 3.1 207.2→72.2 60 35 0.016 0.135 0.003 0.032 0.004 0.035
3.1 207.2→165.4 60 17
Mesosulfuron-methyl 3.0 504.2→182.1 60 31 0.012 0.060 0.003 0.014 0.003 0.014
3.0 504.2→139.2 60 69
Metazachlor 3.3 278.2→134.3 60 27 0.019 0.039 0.012 0.024 0.013 0.026
3.3 278.2→210.3 60 15
Napropamid 3.9 272.2→129.3 60 19 0.003 0.007 0.003 0.006 0.004 0.007
3.9 272.2→171.3 60 21
Quizalofop 3.8 345.2→299.2 60 23 0.520 1.301 2.952 7.381 0.703 1.758
3.8 345.2→273.1 60 25
Thifensulfuron-metyl 2.6 388.1→167.3 66 21 0.442 0.994 0.129 0.324 0.054 0.135
2.6 388.1→141.3 66 21

on peak area. Eight-point calibration curves were constructed mobile phase additives, such as acetic acid and ammonium
using a least-squares linear regression from the injection of acetate at various concentrations was completed. Acetic
sample spiked with solution whose concentrations range from acid in water was preferred in order to enhance ionisation of
1 ng L−1 to 50 μg L−1. pesticide during the electrospray process. The column was
Intra-day and inter-day precision was determined from chosen among a large range of different stationary phases
triplicate analyses of samples spiked with solution at and manufacturers. The optimal separation of the 13
50 μg L−1 during the same day (repeatability) and in five compounds was achieved using 0.0125% (pH=4.04) acetic
successive days (reproducibility). These two parameters acid in aqueous mobile phase and acetonitrile as organic
were expressed as relative standard deviation of result phase on a Zorbax Eclipse C18 (Agilent, 50×2.1 mm,
(RSD%). All the experiments were done in triplicate. 1.8 μm). The retention times for all analytes in the matrix
varied less than 1%. Using the selected conditions,
retention time ranged from 1.6 (carbendazim) to 4.5 min
Results and discussion (diflufenican; Fig. 1 and Table 2).
The optimization of MS parameters was completed by
LC-MS/MS method flow injection analysis (FIA) for each compound. Choice of
the ionization mode, identification of the parent and product
The optimization of the chromatographic separation was ions, and selection of the cone and collision voltages, most
performed from the injection of standard solutions of the favourable for analysis of the target analytes, are shown in
mixture of all analytes. A serial of preliminary experiments, Table 2. For each analyte, the precursor ion was selected by
testing different mobile phases composed of methanol or acquiring in full scan, and the best declustering potential
acetonitrile as an organic phase and water with different was chosen for high production of the parent ion. Then,
Multiresidue method to quantify pesticides in fish muscle 2189

2.4e5
2.3e5 9
2.2e5
2.1e5
2.0e5 1
1.9e5
1.8e5
1.7e5
1.6e5 7
1.5e5
1.4e5 10
Intensity, cps

1.3e5 6
1.2e5 5 8
1.1e5 4 11
1.0e5 13
9.0e4
8.0e4
7.0e4
6.0e4 3
5.0e4
4.0e4
3.0e4 2
2.0e4 12
1.0e4
0.0
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Time, min

Fig. 1 Chromatogram of standard mixture of the 13 pesticides (0.1 μg L−1): 1 carbendazim, 2 fluroxypyr, 3 thifensulfuron-methyl, 4 mesosulfuron-
methyl, 5 isoproturon, 6 metazachlor, 7 dimethachlor, 8 clomazone, 9 axoxystrobin, 10 quizalofop, 11 napropamid, 12 iprodion, 13 diflufenican

collision energies were tested to give the maximum of After the sample homogenization optimisation, the
intensity for the fragment ions. Two different transitions liquid–liquid buffered extraction is investigated. Acetoni-
were monitored per compound when the fragmentation was trile, which gives high extraction recoveries of pesticides in
appropriate: the first and more abundant ion was used for fruits and vegetables, is the most used for its ability to
quantification and the second for confirmation (ratio of separate easily from water with the addition of an
quantity of signal of these two transitions in range of ±10% appropriate mixture of salts. In this study, the Standard
of a standard). Method EN 15662 using acetonitrile and citrate salts was
chosen [35].
Extraction method Moreover, as suggested in the norm EN 15662, a
minimum water percentage is necessary to get good
The QuEChERS protocol contains two major steps: a recoveries in the salting-out extraction. To reach this
salting-out extraction and a dispersive SPE (dSPE) clean- optimum condition, extraction with and without water
up. This dSPE clean-up is not always adapted to the matrix addition was investigated. Table 3 presents the results
and/or analytes studied because it retains the compounds of obtained in both conditions. The best results are obtained
interest or reacts with them. In this case, the second step with the addition of 1 mL of water.
can be cancelled [33] or replaced by a cleaning step by As fish muscles are fat matrices, a clean-up is necessary to
hexane [34] to remove lipids. eliminate lipids. The most used solid phase is PSA, which
In such a protocol, whatever the matrix and the studied should be avoided when analysing fluroxypyr. Moreover, the
molecules, the extraction quality depends on the sample most apolar compounds can be adsorbed on solid phases such
homogenization step. In fact, it should maximize the as C18 and GCB. Consequently, it was chosen to eliminate the
exchange area between sample and extraction solvent. This dispersive SPE step and to remove lipids by the addition of a
step is crucial to ensure excellent recoveries. Thus, fish small volume of hexane [33]. The addition of hexane after
muscles were grinded using an ultra-turrax. Speed of use was sample homogenization led to good pesticides recoveries.
an important parameter: low speed did not crush finely cells When acetonitrile was used without hexane, as described in
and high speed potentially induced losses of pesticides initial QuEChERS method [23], samples were too viscous to
(degradation by too high heat). Middle speed (3/5) was a LC-MS/MS analysis. Few recoveries were obtained (lower
good arrangement if sample tube was maintained at 20 °C in than 50%), and many compounds were not detected such as
a water bath. clomazone, fluroxypyr and iprodion.
2190 A. Lazartigues et al.

Table 3 Comparison between recovery rates obtained with different perch and roach. Consequently, the volume of hexane is
solvents of extraction in the case of spiked perch muscle
not sufficient to purify the extract, leading to lower
Compounds Recovery (%) n=6 recoveries and higher matrix effects (see “Matrix effect”
A B section and Fig. 3). Iprodion was the most impacted
compound with a poor recovery (6%) and a high
Azoxystrobin 80 70
coefficient of variation (25%). Lipid content values for
Carbendazim 76 74
perch and roach tested were similar (1%), and equal
Carbendazim-d4 6 48
recoveries were determined for this two species, except
Clomazone 22 47
for hydrophilic compounds (mesosulfuron-methyl and
Diflufenican 86 83 thifensulfuron-methyl), which did not depend on lipid
Dimethachlor 76 67 content. Consequently, to check and compensate the
Fluroxypyr 72 76 recoveries obtained for the different fish species, internal
Iprodion 48 49 standard was added in the protocol: carbendazim-d4 and
Isoproturon 33 34 isoproturon-d3.
Isoproturon-d3 76 60 To conclude, sample preparation protocol included a
Mesosulfuron-methyl 78 77 sample comminution, a liquid–liquid buffered extraction
Metazachlor 55 50 (according to the EN 15662 version) and a cleaning step
Napropamid 53 40 which was done using hexane and led to satisfactory
Quizalofop 80 70 recoveries (Fig. 2).
Thifensulfuron-methyl 76 74
Matrix effect
A=10 mL acetonitrile then 3 mL hexane; B=10 mL acetonitrile, then
3 mL hexane and 1 mL water
Matrix effects like enhancement or suppression of signals
can severely compromise the quantitative analysis of
The extraction optimisation was performed on perch environmental samples. In order to evaluate the influ-
muscles and extended then to carp and roach to apply ence of the matrix on the analysis, samples were spiked
method on fatty matrices. Results on performance of after extraction, and the areas (Afish) were compared
extraction for the three species are presented in Fig. 2. with those of injected standards (As) at 50 μg/L. The
Method obtained generally good recoveries for all matrix effect was calculated by means of the following
compounds despite the complexity of the matrices. equation:
However, recoveries of most of the compounds are  
significantly lower for carp compared to roach and perch. ðAs  Afish Þ
Suppression=enhancement signalð%Þ ¼  100
Indeed, lipid content in carp is five times higher than in As

Fig. 2 Comparison of
recoveries (percent) for the
three species tested
(roach, perch and carp)
Multiresidue method to quantify pesticides in fish muscle 2191

Matrix effects are shown in Fig. 3. The percentage varies to-noise ratio (S/N) was >3 for both transitions and RSD
from −12% to +90% for roach, from 17% to +90% for of the respective ion ratios was within the tolerances
perch and from −69% to +98% for carp (Fig. 3). RSD for specified in the EU legislation.
matrix effect are all inferior than 20%. These results
indicate that strong MS signal suppression effects were Method application
observed for most of the compounds except iprodion. Thus,
to compensate this matrix effect and avoid any under/over Method was applied on 50 samples of muscle of roaches
estimation of pesticides, a matrix-matched calibration in caught in five dam ponds. Only few of these samples were
eight points was used. contaminated; molecules were generally not detected
(<LOD). Carbendazim, diflufenican and isoproturon were
Method validation often detected (<LOQ) or even quantified; maxima values
were 0.12 and 0.64 ng g−1 for carbendazim and diflufenican,
The performance of the method was evaluated in term of respectively (Fig. 4). Comparison with literature was difficult
specificity, precision (intra-day precision and inter-day because, to our knowledge, there is a lack of data about
precision) and linearity; recoveries and limits of detection levels of these compounds in fish matrices. In conse-
and quantification have already been discussed previously. quence, we checked the method analyzing 104 supple-
Specificity was assessed by the analysis of three blank mentary samples. They were muscles of fish (carp and
fish samples extracted by the optimised QuEChERS perch) orally contaminated (500 ng g−1 body weight via
method. No interference was detected at the analyte contaminated pellets) by a mixture of the 13 pesticides
retention time. Reproducibility and repeatability expressed during 6 days and depurated during 12 days in clean water.
as relative standard deviation, RSD%, were lower for the Maximal coefficient variation between repetitions (for a
three different fish species than 28% for intra-day and same sampling time) was 20% (experimental and analyt-
29% for inter-day analysis, respectively. Eight-point ical conditions coupled). Residue concentrations ranged
calibration curves gave very good fits, r2 >0.99, over from <LOD before exposure phase to maximal values
the established concentration range from LOQ to during exposure phase (Fig. 5). This experience confirmed
81 ng g−1 for all the compounds. Finally, in real sample, that the method developed could quantify compounds
results were regarded as positively identified when signal- incorporated in fish muscle.

Fig. 3 Evaluation of
matrix effect
2192 A. Lazartigues et al.

Fig. 4 Frequency of
contaminated muscles of
roach in dam ponds (percent)

Conclusion by liquid chromatography and MS. Good sensitivity and

recoveries were generally observed in spite of strong matrix
This work proposed an easy and quick method using few effects. Quantification limits were below 1 ng g−1 except for
solvent quantities to quantify 13 pesticides in fish muscle clomazone (1.8 ng g−1) and quizalofop (7.4 ng g−1). Low
(azoxystrobin, clomazone, diflufenican, dimethachlor, car- quantification limits are often needed in environmental and
bendazim, iprodion, isoproturon, mesosulfuron-methyl, met- health research. For example, they were necessary to
azachlor, napropamid, quizalofop and thifensulfuron-methyl). evaluate behaviour or impact of contaminants, particularly
Method was based on an ultra-turrax homogenisation coupled in the case of low doses and/or long time exposures to a
to an adapted QuEChERS extraction. Analysis was performed mixture of pesticides.

Fig. 5 Maximal concentrations

(microgrammes per
kilogramme) quantified
during an oral exposure of carp
and perch
Multiresidue method to quantify pesticides in fish muscle 2193

Acknowledgments This work was funded by the Agence de l'Eau determination of emerging pollutants in water by solid-phase
Rhin-Meuse and the Zone Atelier Moselle, which are gratefully extraction and liquid chromatography-tandem mass spectrometry.
acknowledged. J Chromatogr A 1216:2958–2969
19. Barrek S, Cren-Olivé C, Wiest L, Baudot R, Arnaudguilhem C,
Grenier-Loustalot MF (2009) Multi-residue analysis and ultra-
trace quantification of 36 priority substances from the European
References
Water Framework Directive by GC-MS and LC-FLD-MS/MS in
surface waters. Talanta 79:712–722
1. Reemtsma T, Jekel M (2006) Organic pollutants in the water 20. Baugros JB, Giroud B, Dessalces G, Grenier-Loustalot MF, Cren-
cycle. Wiley-VCH Verlag GmBH & Co. KGaA, Weinheim Olivé C (2008) Multiresidue analytical methods for the ultra-trace
2. Loos R, Gawlik BM, Locoro G, Rimaviciute E, Contini S, Bidoglio quantification of 33 priority substances present in the list of
G (2009) EU-wide survey of polar organic persistent pollutants in REACH in real water samples. Anal Chim Acta 607:191–203
European river waters. Environ Pollut 157:561– 568 21. Pico Y, Barcelo D (2008) The expanding role of LC-MS in
3. Devault DA, Gérino M, Laplanche C, Julien F, Winterton P, analyzing metabolites and degradation products of food contam-
Merlina G, Delmas F, Lim P, Miguel Sánchez-Pérez J, Pinelli E inants. Trac Trends Anal Chem 27:821–835
(2009) Herbicide accumulation and evolution in reservoir sedi- 22. Blasco C, Font G, Pico Y (2005) Analysis of pesticides in fruits
ments. Sci Total Environ 407:2659–2665 by pressurized liquid extraction and liquid chromatography-ion
4. Bony S, Gaillard I, Devaux A (2010) Genotoxicity assessment of trap-triple stage mass spectrometry. J Chromatogr A 1098:37–43
two vineyard pesticides in zebrafish. Int J Environ Anal Chem 23. Anastassiades M, Lehotay SJ, Stajnbaher D, Schenck FJ (2003) Fast
90:421–428 and easy multiresidue method employing acetonitrile extraction/
5. Marchand J, Tanguy A, Charrier G, Quiniou L, Plee-Gauthier E, partitioning and “dispersive solid phase extraction” for the determi-
Laroche J (2006) Molecular identification and expression of nation of pesticides residues in produce. J AOAC Int 86:412–431
differentially regulated genes of the European flounder, Platichthys 24. Wilkowska A, Biziuk M (2011) Determination of pesticide
flesus, submitted to pesticide exposure. Mar Biotechnol 8:275–294 residues in food matrices using QuEChERS methodology. Food
6. Gorbatiuk LO (2010) Physiological-biochemical response of fish Chem 125:803–812
to the pesticide effect (a review). Hydrobiol J 46:78–87 25. Kolberg DI, Prestes OD, Adaime MB, Zanella R (2011) Develop-
7. Hu G, Dai J, Mai B, Luo X, Cao H, Wang J, Li F, Xu M (2010) ment of a fast multiresidue method for the determination of pesticides
Concentrations and accumulation features of organochlorine in dry samples (wheat grains, flour and bran) using QuEChERS
pesticides in the Baiyangdian lake freshwater food web of North based method and GC-MS. Food Chem 125:1426–1442
China. Arch Environ Contam Toxicol 58:700–710 26. Chen G, Cao P, Liu R (2011) A multi-residue method for fast
8. John PJ, Prakash A (2003) Bioaccumulation of pesticides on some determination of pesticides in tea by ultra performance liquid
organs of freshwater catfish Mystus vittatus. Bull Environ Contam chromatography-electrospray tandem mass spectrometry com-
Toxicol 70:1013–1016 bined with modified QuEChERS sample preparation procedure.
9. Sun F, Wong SS, Li GC, Chen SN (2006) A preliminary Food Chem 125:1406–1411
assessment of consumer's exposure to pesticide residues in 27. Gonzalez-Curbelo MA, Hernandez-Borges J, Ravelo-Perez LM
fisheries products. Chemosphere 62:674–680 (2011) Insecticides extraction from banana leaves using modified
10. Malik AK, Blasco C, Picó Y (2010) Liquid chromatography-mass QuEChERS method. Food Chem 125:1083–1090
spectrometry in food safety. J Chromatogr A 1217:4018–4040 28. Rawn DFK, Judge J, Roscoe V (2010) Application of the
11. Reemtsma TJ (2003) Liquid chromatography-mass spectrometry QuEChERS mehod for the analysis of pyrethrins and pyrethroids
and strategies for trace-level analysis of polar organic pollutants. J in fish tissues. Anal Bioanal Chem 397:2525–2531
Chromatogr A 1000:477–501 29. Lehotay SJ, Mastovska K, Yun SJ (2005) Evaluation of two fast
12. Pico Y, Blasco C, Font G (2004) Environmental and food and easy methods for pesticide residue analysis in fatty food
applications of LC-tandem mass spectrometry in pesticide- matrices. J AOAC Int 88:630–638
residue analysis: an overview. Mass Spectrom Rev 23:45–85 30. Couteux A (2008) Index phytosanitaire. ACTA, Lille
13. Alder L, Greulich K, Kempe G, Vieth B (2006) Residue analysis 31. Pang G-F, Cao Y-Z, Fan C-L, Jia G-Q, Zhang J-J, Li X-M, Liu Y-M,
of 500 high priority pesticides: better by GC-MS or LC-MS/MS? Shi Y-Q, Li Z-Y, Zheng F, Lian Y-J (2009) Analysis method study on
Mass Spectrom Rev 25:838–865 839 pesticide and chemical contaminant multiresidues in animal
14. Kuster M, Lopez de Alda MJ, Barcelo D (2006) Analysis of muscles by gel permeation chromatography cleanup, GC/MS, and
pesticides in water by liquid chromatography-tandem mass LC/MS/MS. JAOAC Int 92:933–940
spectrometric techniques. Mass Spectrom Rev 25:900–916 32. Council of European Communities (1986) Council Directive 86/609/
15. Ferrer I, Thurman EM, Zweigenbaum JA (2007) Screening and EEC of 24 November 1986 on the approximation of laws, regulations
confirmation of 100 pesticides in food samples by liquid and administrative provisions of the Member States regarding the
chromatography/tandem mass spectrometry. Rapid Commun Mass protection of animals used for experimental and other scientific
Spectrom 21:3869–3882 purposes. Available at http://europa.eu/legislation_summaries/
16. Nardelli V, Dell'Oro D, Palermo C, Centonze D (2010) Multi- environment/nature_and_biodiversity/l28104_en.htm
residue method for the determination of organochlorine pesticides 33. Aguilera-Luiz MM, Martinez Vidal JL, Romero-Gonzales R,
in fish feed based on a cleanup approach followed by gas Garrido Frenich A (2008) Multi-residue determination of veteri-
chromatography-triple quadrupole tandem mass spectrometry. J nary drugs in milk by ultra-high-pessure liquid chromatography-
Chromatogr A 1217:4996–5003 tandem mass spectrometry. J Chromatogr A 1205:10–16
17. Petrovic M, Farré M, De Alda ML, Perez S, Postigo C, Köck M, 34. Przybylski C, Segard C (2009) Method for routine screening of
Radjenovic J, Gros M, Barcelo D (2010) Recent trends in the pesticides and metabolites in meat based baby-food using extraction
liquid chromatography-mass spectrometry analysis of organic and gas chromatography-mass spectrometry. J Sep Sci 32:1858–1867
contaminants in environmental samples. J Chromatogr A 35. NF V03-061; NF EN 15662:2009-01-01 Foods of plant origin—
1217:4004–4017 determination of pesticide residues using GC-MS and/or LC-MS/
18. Rodil R, Quintana JB, López-Mahía P, Muniategui-Lorenzo S, MS following acetonitrile extraction/partitioning and clean-up by
Prada-Rodríguez D (2009) Multi-residue analytical method for the dispersive SPE-QuEChERS-method