Analytica Chimica Acta 464 (2002) 303–312

Supercritical fluid extraction of organochlorine pesticides

and some metabolites in frogs from National Park
of Ordesa and Monte Perdido
C. Nerı́n a,∗ , R. Batlle a , M. Sartaguda a , C. Pedrocchi b
a Departamento de Quı́mica Analı́tica, Centro Politécnico Superior de Ingenieros, Universidad de Zaragoza,
C/Ma de Luna 3, E-50018 Zaragoza, Spain
b Instituto Pirenaico de Ecologı́a, CSIC, Avda Regimiento de Galicia s/n, E-22700 Jaca (Huesca), Spain

Received 18 February 2002; received in revised form 16 May 2002; accepted 16 May 2002

Abstract
The optimization of supercritical fluid extraction (SFE) with CO2 for the extraction of a series of organochlorine pesticides
(OCPs) and some metabolites in biota samples (frogs) from the National Park of Ordesa and Monte Perdido (Spain) and
further analysis by GC–electron capture detector (ECD) is carried out. An experimental design is applied to the optimization
of SFE. The Soxhlet extraction of the same samples is optimized too, together with the necessary clean-up step using four
different solid cartridges. The analytical procedures as well as the results obtained are compared. Seven in sixteen OCPs
under study showed recovery values above 80%, when using Soxhlet while nine OCPs were over 80% in the case of SFE. The
analysis of frog samples showed values below 7 ng/g on the basis of dry matter for all the OCPs under study. The analytical
procedures as well as the results obtained in real samples are discussed.
© 2002 Elsevier Science B.V. All rights reserved.
Keywords: Organochlorine pesticides; Metabolites; Analysis; Biota; SFE

1. Introduction ecosystems in which they are introduced [1]. OCPs are

accumulative and persistent compounds in both biota
Since 1940, organochlorine pesticides (OCPs) have and the environment and as liposoluble compounds
been applied to protect crops against the pests, allow- they are difficult to be metabolized by living beings
ing a spectacular increase of production as well as the [2]. It has been also demonstrated that some species
erradication of illness such as malaria and typhus. The have developed new protection mechanisms which
use of pesticides has been assumed as something nec- make them resistant to further applications of OCPs
essary and it demands well established rules to prevent although such application affects the trophic chain [3].
the increasing contamination. Although the restric- When an OCP is applied, it is distributed into
tions that most of the pesticides have around the world, the different environmental compartments such as
OCPs have caused the lost of equilibrium of those the atmosphere, water, soil and biota, depending
on the physical conditions. For all these reasons,
∗ Corresponding author. Tel.: +34-976761873; most of OCPs are now banned in most of the devel-
fax: +34-976761861. oped countries. However, developing countries still
E-mail address: [email protected] (C. Nerı́n). use some OCPs and as a result of the atmospheric

0003-2670/02/$ – see front matter © 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 0 0 3 - 2 6 7 0 ( 0 2 ) 0 0 4 3 4 - 8
304 C. Nerı́n et al. / Analytica Chimica Acta 464 (2002) 303–312

transport, these pollutants are found everywhere, 2. Experimental

even in protected areas. This is the case of National
Park of Ordesa and Monte Perdido, a natural protected 2.1. Reagents and solutions
and beautiful area of Spanish Pyrenees, far from both
agricultural and industrial uses, in which some pol- Heptachlor, heptachloroepoxide, ␣-endosulphan
luted living species were found in 1993 [4]. Previous and endosulphan-sulphate were from Riedel-de Häen
studies in the area showed an atmospheric transport (Seelze, Germany). ␣-Hexachlorocyclohexane (HCH),
which through the rain, wind and snow [5,6], carried ␤-hexachlorocyclohexane, ␥-hexachlorocyclohexane
some pesticides to this area. (lindane), dieldrin, endrin, aldrin, endrin-aldehyde,
In order to study the influence of this transport ␤-endosulphan, p,p -DDD, p,p -DDE, p,p -DDT were
on the trophic chain, a research work was designed from Dr. Ehrenstorfer (Augsburg, Austria). PCB-52
which involves the determination of the concentra- from Chem Service (West Chester, USA) was used as
tion of several organochlorine compounds and some internal standard.
metabolites in the trophic chain starting with a kind Dichloromethane, n-hexane, isopropanol and
of grass-eater, such as the locust Orthopter orthopter toluene of residue analysis quality and silica gel 60
[7], and then frogs, birds and mammals living in the (70–230 mesh ASM) for chromatography were from
area. The present paper shows the study carried out in Merck (Merck, Darmstadt, Germany). Diethyl-ether
frogs. and anhydrous sodium sulphate for analysis quality
As very low concentrations are expected, their were from Panreac (Panreac Quı́mica, Barcelona,
analysis required a very sensitive and accurate proce- Spain). Florisil (60–100 mesh) was from Fluka.
dure. A quite extensive literature about the analytical
procedures for OCPs has been reviewed but none of 2.2. Selection of the pesticides
the papers deal with a sample similar to frogs. On the
other hand, the analytical procedures usually involve Previous studies carried out on wild mammals in
an extraction step with solvent, a clean-up system the same National Park [4] showed the presence of
and a final concentration step of the extract, previous OCPs residues in serum. Also the analysis of snow,
to the GC–electron capture detector (ECD) analysis. air and water as well as of locust captured in the area
In the last few years, the trends for simplifying the revealed that the concentration of pesticides increased
analytical procedures have driven to the development from snow and water to biota. The appearance in the
of new sample treatment techniques which avoid the later samples of some metabolites of organochlorine
clean-up step. In this context, supercritical fluid ex- pesticides suggested the inclusion of these compounds
traction (SFE) appears as one of the most appropriate in the list of pollutants under study. For this reason,
technique for this type of analysis in different solid the following compounds were considered: ␣-HCH,
matrices [8–11]. It allows to do selective extractions ␤-HCH, ␥-HCH (lindane), ␦-HCH, heptachloroepox-
of different chemicals without additional clean-up ide, heptachlor, aldrin, dieldrin, endrin, endrin-alde-
steps as well as to use little sample amounts which hyde, endosulphan-sulphate, ␣-endosulphan, ␤-endo-
in this case, working with small animals or protected sulphan, p,p -DDT, p,p -DDE and p,p -DDD.
species, is a very important feature. In contrast, Soxh-
let extraction has been widely used for OCPs [12–16]. 2.3. Samples
It requires high volume of organic solvents and a high
handling degree and time consuming. The frogs were captured in 1996 in the Pirineo
This paper shows the SFE optimization for the Aragonés (Spain). Two different groups of frogs were
determination “off-line” of some OCPs and their available. The first one corresponded to those cap-
metabolites by GC–ECD in frogs, as a potential tured far from the Parque Nacional de Ordesa y Monte
bioindicators of the pollution. The analytical proce- Perdido and they were used as blank samples. Spiked
dure as well as the results of OCPs in frogs from the samples were prepared from these blank samples for
Parque Nacional de Ordesa y Monte Perdido (Spain) optimization. The second group was captured in the
are discussed. mentioned National Park. All of them belong the
 C. Nerı́n et al. / Analytica Chimica Acta 464 (2002) 303–312 305

Fig. 1. Chromatogram of a standard solution of OCPs in isopropanol by GC–ECD. Compounds are identified in Table 1.

autoctonous specie Rana pirenaica of a high ecologi- copolymer (Sugelabor, Madrid, Spain). The program
cal value and for this reason a special permission was used was as follows: initial temperature 50 ◦ C, hold for
necessary. 2 min, rate at 25 ◦ C/min till 180 ◦ C and then hold for
Once in the laboratory, the samples were grinded to 15 min, rate at 25 ◦ C/min till 290 ◦ C and then hold for
make an homogeneous paste and frozen for storage at 2 min. Detector temperature was 300 ◦ C. Carrier gas
20 ◦ C below cero. Skin and bones as well as the head used was hydrogen-C50 (Carburos Metálicos, Spain)
were included in the mixture to avoid the use of a at 1 ml/min and N2 at 30 ml/min was used as make up
higher number of animals. The samples were defrozen gas. Star Chromatography Workstation 4.51 version
just before the analysis. The analysis were performed (Varian, Texas, USA) and a personal computer were
in 1997. According to the literature [3], no degrada- used to obtain the data. Fig. 1 shows a chromatogram
tion was expected of the residue pesticides during the with the retention time of the compounds under study.
storage time and conditions.

2.4. Apparatus 3. Soxhlet extraction

Soxhlet and thermostatic bath (Precis Pat S-48200) Soxhlet extraction with dichloromethane as extract-
from Selecta (Selecta, Abrera, Spain) were used. ing solvent was applied for 8 h at 40 ◦ C. The cycling
Supercritical fluid extractor was from Varian Star rate was about 12 cycles per hour. The solid sample
SFE and it was equipped with a Premaster Star SFE was mixed with anhydrous sodium sulphate at the rate
extractor, a Modifier Pump Star SFE and the Accu- 1:5 (w/w) and it was introduced into a cellulose car-
trap Star SFE collector (Suprex Corporation, Pitts- tridge covered by silanized glass wool. After 8 h, the
burg, USA). Gas chromatograph Varian Star 3400 Cx extract was cleaned-up in a minicolumn containing
equipped with a capillar split-splitless injector as well 2 g of 3% deactivated silica gel, which was then eluted
as a septum equipped programmable injector (SPI), with 40 ml of hexane. The final extract was concen-
an autosampler Varian 8200 Cx and electron capture trated to about 1 g of solution and filtered through a
detector (ECD) of 63 Ni was used. The column used syringe filter of 0.22 ␮m of pore size. PCB-52 was
was SGL-5 60 m × 0.25 mm i.d. and 0.25 ␮m of film added as internal standard before the GC injection.
thickness of 5% diphenyl, 95% dimethylpolysiloxane All the solutions were gravimetrically controlled.
306 C. Nerı́n et al. / Analytica Chimica Acta 464 (2002) 303–312

4. SFE (B) a 3% deactivated silica (2 g) eluted with 15 ml of

n-hexane:diethylether (70:30 v/v) mixture;
The extraction cell of 3 ml was filled with the mix- (C) a 5% deactivated florisil (2 g) eluted as mentioned
ture of sample and anhydrous sodium sulphate (1:5) in B;
in sandwich mode using silanized glass wool at both (D) a 3% deactivated silica (2 g) eluted with 40 ml of
the bottom and the top of the cell to protect the cell n-hexane.
seals. A pressure of 425 atm and 35 ◦ C of temperature
with a flow of 2 ml/min of CO2 for 5 min of static In all cases the adsorbents were previously activated
extraction and then 10 min of dynamic extraction at 450 ◦ C for 5 h and further deactivated with distilled
were applied to the extraction module. The adsorption water. They were stored in a desicator 24 h before
temperature in the collection module (Accutrap) was using. This procedure was repeated weekly.
−15 ◦ C and the desorption temperature was 30 ◦ C The recovery studies were carried out by spiking
using 1.7 ml of n-hexane as eluting solvent. The final the adsorbent with 300 ␮l of a solution containing
weight of the extract was considered instead of the 800 ng/g of each pesticide and then applying the elu-
volume in order to be consistent with the gravimetric tion system and analysis before described. The results
control, according to the Quality Control require- obtained are shown in Table 1. Both the recovery val-
ments. PCB-52, used as internal standard, was added ues and the chromatogram obtained in each case were
to the final n-hexane extract before the GC–ECD taken into account for the final selection. It was ob-
analysis. served that the presence of diethyl ether in the elu-
tion solvent carried a higher amount of fat molecules
as well as other molecules which were co-eluted with
5. Results and discussion the compounds under study and interfered the anal-
ysis. For this reason, the use of diethyl ether is not
5.1. Soxhlet extraction and clean-up recommended even though the recovery values were
very high.
The main variables affecting the Soxhlet extraction
As can be seen in Table 1, ␤-HCH and p,p -DDE
are the extraction time, the temperature and the ex-
gave values higher that 100% in all cases. This
traction solvent. According to the previous studies [7]
fact could be attributed to several causes. First, the
dichloromethane at 40 ◦ C for 8 h was shown to be the
poor separation of the pairs ␤- and ␥-HCH and also
best extraction agent for these analytes in biota sam-
p,p -DDE and dieldrin in the chromatographic column
ples. However, in these conditions, fat molecules as
used. This would provide inadequate quantitative re-
well as other interferences present in the samples are
sults. The second reason could be the decomposition
co-extracted and a clean-up step is necessary. The use
of parent compounds in the case of p,p -DDE in the
of a solid bed of an appropriate adsorbent such as sil-
injection port, as other authors mention. Although
ica, alumina or florisil or even a combination of several
the different studies carried out changing the injec-
adsorbents in a minicolumn, which is further eluted
tor port and improving the analytical conditions, no
with a solvent, has been widely proposed in the liter-
clear evidence was found to justify the high results
ature [17–20].
of p,p -DDE. Based on these experimental data, the
As the analytes under study cover a wide range of
selected clean-up system was D.
polarity, several adsorbents as well as different com-
binations of them were tested, each one having a dif-
5.2. Recovery study using the Soxhlet extraction
ferent elution solvent in both volume and nature. The
following systems were considered:
Several frog samples were spiked with 300 ␮l of a
(A) a 3% deactivated silica (2 g) + 5% deactivated standard solution containing 800 ng/g of each OCP
florisil (2 g) using two sequential steps of elution, under study and 24 h after being fortified the samples
the first one with 20 ml of n-hexane and the second were analyzed following the optimum procedure de-
one with 20 ml of n-hexane:diethylether (70:30 scribed under the experimental section. The recovery
v/v) mixture; values as well as the RSD expressed as (%) are shown
 C. Nerı́n et al. / Analytica Chimica Acta 464 (2002) 303–312 307

Table 1
Recovery values for some OCPs using Soxhlet extraction and different clean-up procedures
Peak Compound tR (min) 3% Silica–5% 3% Silica–hexane 5% Florisil–hexane 3% Silica–hexane
no. florisil–hexane ether ether ether
1 ␣-HCH 17.24 98.3 ± 26 97.1 ± 29 92.1 ± 23 99.2 ± 20
2 ␤-HCH 19.28 135.8 ± 11 112.5 ± 26 120.9 ± 5.9 102.8 ± 13
3 ␥-HCH 19.57 93.4 ± 32 99.5 ± 28 91.0 ± 15 97.6 ± 21
4 ␦-HCH 21.74 86.3 ± 31 97.8 ± 26 83.8 ± 13 99.0 ± 12
5 Heptachlor 25.08 95.0 ± 50 98.3 ± 19 91.8 ± 41 99.8 ± 28
6 PCB-52 26.07
7 Aldrin 26.78 82.4 ± 27 88.5 ± 14 82.8 ± 20 93.1 ± 12
8 Heptachloroepoxide 28.78 90.3 ± 24 96.7 ± 8.3 89.7 ± 18 93.8 ± 5.9
9 ␣-Endosulphan 30.44 90.6 ± 31 97.7 ± 8.0 87.7 ± 27 86.5 ± 11
10 p,p -DDE 31.80 120.9 ± 2 132.1 ± 14 108.5 ± 16 121.4 ± 8.3
11 Dieldrin 31.90 85.3 ± 27 93.3 ± 10 88.5 ± 13 90.3 ± 18
12 Endrin 33.50 90.2 ± 30 98.2 ± 8.6 90.9 ± 18 92.3 ± 15
13 ␤-Endosulphan 33.85 66.7 ± 17 74.3 ± 18 67.6 ± 9.4 99.1 ± 7.6
14 p,p -DDD 34.54 85.9 ± 34 95.6 ± 19 90.3 ± 11 96.1 ± 10
15 Endrin-aldehyde 35.11 90.5 ± 42 99.7 ± 26 88.5 ± 7.0 74.1 ± 16
16 Endosulphan-sulphate 36.65 84.5 ± 45 94.6 ± 29 90.0 ± 21 98.3 ± 26
17 p,p -DDT 36.88 87.5 ± 39 97.0 ± 21 86.4 ± 30 99.1 ± 27
x ± f σ (n = 3; f = 4.3; 95% confidence limit).

in Fig. 2. As can be seen, the recovery of some com- with the frog sample occurs. On the other hand,
pounds is below 70% and even lower than 30% for p,p -DDE recovery is considerably higher than 100%
some of them, such as dieldrin, endrin, ␤-endosulphan, while dieldrin recovery is very low, which could
endrin-aldehyde and endosulphan-sulphate, which be attributed to either the poor separation between
means that a clear interaction of the added compounds them, as was earlier mentioned, or to the presence

Fig. 2. Recovery studies by Soxhlet procedure.

308 C. Nerı́n et al. / Analytica Chimica Acta 464 (2002) 303–312

of p,p -DDE in the blank sample. Direct analysis of Table 3

blank samples was always below the detection limit, Limit values applied to the experimental design for optimizing the
extraction step in SFE
so this attempt could not be confirmed. These results
show that the sample matrix strongly affects the be- Variable Minimum Maximum
haviour of these OCPs in the extraction and clean-up Pressure (atm) 200 425
procedures used. Other authors [19,20] found simi- Temperature (◦ C) 35 75
lar results in biota samples, which means that other Flow of CO2 (ml/min) 2 4
extraction procedure is required. Conditions of the collection step: adsorption Ta : −15 ◦ C; des-
orption Ta : 30 ◦ C; volume of hexane (elution solvent): 1.7 ml; t
5.3. Supercritical fluid extraction (SFE); (static): 5 min; t (dynamic): 10 min.
optimization based on an experimental design
CO2 flow during the dynamic extraction. The experi-
The variables involved in the SFE can be classi- mental design used, based on previous studies [8,21]
fied in two groups: (a) those affecting the extraction is shown in Table 3. For this study, the optimum values
step and (b) those related to the collection of the an- for the collection step optimized earlier were applied,
alytes. Both groups of variables were independently but the sample preparation was different.
optimized using a factorial design in which all the Some frog samples were spiked with 80 ␮l of a stan-
likely combinations of factors affecting the experi- dard solution of pesticides containing 800 ng/g of each
ment, including the crossed interaction of variables, compound. The 0.5 g of this spiked sample were mixed
are considered. with anhydrous sodium sulphate (1:5 w/w) and then
To optimize the collection step, the recommenda- the 3 ml extraction cell was filled, adding silanized
tions from Snyder et al. [8] were taken into account, glass wool on the bottom and on the top. The same cri-
and the extreme values used are shown in Table 2. terium as that mentioned in the previous studies was
Each experiment was carried out as follows: The ex- used for optimization. The optimum values obtained
traction cell was filled with silanized glass wool on the were the following: 2 ml/min of CO2 flow, 425 atm
bottom and then anhydrous sodium sulphate to which and 35 ◦ C in the extraction cell. Table 4 lists the opti-
500 ␮l of a standard solution containing 800 ng/g of mum values found
each pesticide was added. The percentage of total re-
covery of pesticides was considered as criterium of 5.3.1. Recovery studies by SFE–GC–ECD
optimization. The only relevant variable found in the Fig. 3 shows the plot of recoveries obtained. As
collection step, after the statistical evaluation of the can be seen, all the values are higher than 40%, be-
data, was the volume of solvent for elution, which was ing the behaviour of dieldrin, endrin, ␤-endosulphan,
1.7 ml. The other two variables were no significant and endrin-aldehyde and endosluphan-sulphate much bet-
for this reason, the middle point was adopted in each ter than in the Soxhlet extraction. On the other hand,
one. p,p -DDE recovery is also higher than 100% as it was
In the extraction step, the following variables were in Soxhlet extraction, which could be attributed to
optimized: the pressure and temperature in the cell and the presence of this compound in the blank sample.
The direct analysis of this compound in the sample
Table 2 Table 4
Limit values applied to the experimental design for optimizing the Optimum values found for the extraction of OCPs from frog
collection step in SFE samples by SFE
Variable Minimum Maximum Extraction Collection
Adsorption Ta (◦ C) −30 0 Pressure (atm) 425 Adsorption temperature (◦ C) −15
Desorption Ta (◦ C) 15 45 Temperature (◦ C) 35 Desorption temperature (◦ C) 30
Elution volume (hexane) 0.7 1.7 Flow of CO2 2 Volume of solvent (ml) 1.7
Extraction conditions: P: 350 atm; T: 75 ◦ C, t (static): 5 min; t (ml/min)
(dynamic): 10 min. t (static): 5 min; t (dynamic): 10 min.
 C. Nerı́n et al. / Analytica Chimica Acta 464 (2002) 303–312 309

Fig. 3. Recovery studies by SFE procedure.

gave values under the detection limit and for this rea- using an homogeneous mixture of the sample and the
son, this attempt could not be confirmed. The ob- drier which is essential to get accurate values. How-
tained differences between Soxhlet and SFE are not ever, Echarri et al. [25] and Wells and Echarri [26]
only due to the extraction itself, but SFE does not re- worked with anhydrous sodium sulphate as drier agent
quire a further clean-up step and it is well known that for biota samples and no adsorption effect was ob-
the clean-up steps are usually the most likely sources served. Also the liofilization process before the extrac-
of error in the quantitative determination of pesticides tion has been proposed as a good approach.
[22]. The main problem of the presence of water is its
solubility in the supercritical CO2 which is approxi-
5.4. Comparison of Soxhlet and SFE techniques mately 0.3%. This fact not only affects the extraction
itself but the ice which is formed in the restrictor can
Recovery efficiency is an important factor when an block the outlet and consequently the CO2 flow is not
analytical method is evaluated. It can be considered constant. Also in this case the water drops in the final
that a method is acceptable when the recoveries are extract can be demixed producing two phases, which
over 80% [17]. Figs. 2 and 3 clearly show that SFE do not provide quantitative values [27].
provides a higher recovery in all cases with the only ␤-HCH and p,p -DDE are always recovered over
exception of p,p -DDT whose values are 77% in Soxh- 100%. As was mentioned earlier, in the case of
let extraction and 47.3% in SFE. Seven compounds in p,p -DDE it could be attributed to a poor separation
16 gave values higher than 80% in Soxhlet extraction of dieldrin, whose recovery is always quite poor.
whereas nine compounds were recovered over the rec- p,p -DDE is also a metabolite from DDT and other
ommended value in SFE, being only three compounds authors have described the decomposition of DDT
with values around 50%. The explanation for this low compounds in the injection port at high temperature.
values is not very clear. Reimer et al. [23] described This is the reason why on column cold injectors are
that the mixture of the samples with the drier agent preferred for this type of analysis. In fact, DDT re-
such as Na2 SO4 decreases the recovery of organochlo- coveries are lower than 100% in both Soxhlet and
rine compounds and they suggest to introduce both SFE procedures. Unfortunately, on column injection
the drier agent and the sample in sandwhich beds in- couldn’t be used in this study. On the other hand, no
stead of mixing them. This fact is specially critic in correlation was found between the compound proper-
␣- and ␤-endosulphan and could be the explanation ties (structure, volatility, etc.) and recovery by either
in this case. In contrast, Valverde et al. [24] insist of method.
310 C. Nerı́n et al. / Analytica Chimica Acta 464 (2002) 303–312

Another factor in favour with the SFE extraction 5.5. Analysis of pesticides in frogs
is its better precision which reduces the number and
magnitude of errors. This is again because the SFE Four different samples of frogs “Rana pirenaica”
procedure is more automatic and has a lower number captured in the National Park of Ordesa and Monte
of steps. Figs. 2 and 3 show the RSD values and in Perdido (Spain) were analyzed. The direct analysis
general those corresponding to SFE are lower than using an internal standard showed that the real val-
in Soxhlet. The variation coefficient is quite high for ues were below the quantification limit (between 7.39
p,p -DDE and dieldrin in SFE and this fact could be and 30.28 ng/g) and for this reason, the standard addi-
attributed to the chromatographic separation. tion procedure was applied. Six different concentration
Concerning the environmental factors, SFE is much values from 0 to 202.24 ng/g of pesticides were used
more convenient than Soxhlet, as the former does not and in all cases the gravimetric control was applied
require the use of organic solvents (only 1.7 ml of to the solutions. The results obtained are shown in
hexane) while the later employs 210 ml of them, and Fig. 4.
among them 150 ml are of dichloromethane, a chlori- It is interesting to point out that all the pesticides and
nated solvent, expensive and toxic. metabolites were found in the four samples at very low
The comparison of total time of the analytical pro- concentration level, with the only exception of ␦-HCH
cedure, a more and more important factor, shows that which was no present in two of the samples. The trends
SFE is considerably faster than Soxhlet extraction, as of concentration values are similar in all samples,
high volumes of organic solvents are handled and a which show higher concentration of the most persis-
clean-up step is necessary when using Soxhlet while tent compounds such as ␤-HCH and p,p -DDT [28].
it does not in SFE. On the other hand, it can be observed that metabo-
SFE is also environmental friendly whereas Soxhlet lites such as p,p -DDD, p,p -DDE, endrin-aldehyde,
is not, due to the high consumption of organic solvents. endosulphan-sulphate and heptachlor are present at

Fig. 4. Concentration of OCPs in frog samples from National Park of Ordesa and Monte Perdido.
 C. Nerı́n et al. / Analytica Chimica Acta 464 (2002) 303–312 311

Acknowledgements

This work has been financed by the Project PMA

1993–1996, PMA 1996–1999 and PMA 2001 from
CONAI, Gobierno de Aragón, Spain.

References

[1] R.J. Cremlym, Agrochemicals: Preparation and Mode of

Fig. 5. Concentration of OCPs in some biota samples of the trophic Action, Wiley, New York, 1991.
chain from National Park of Ordesa and Monte Perdido. [2] R. Garrido-Lesteache, M.E. Garrido-Lesteache, Toxicologı́a
de pesticidas, Rafael Garrido-Lesteache Cabrera, Madrid,
Spain, 1988.
higher concentration than their parents, which con- [3] X. Duran, L.A. Fernández, L. Reales, La omnipresencia de
firms a strong interaction of these pesticides with los Organoclorados, Departament de Medi Ambient de la
biota. Generalitat de Catalunya, Barcelona, 1992.
These results agree with those obtained in previ- [4] M.C. Martı́nez, S. Gargallo, P. Ferrando, J. Maluenda, D.
Raldua, J. Quı́lez, J. Gonzalo, C. Pedrocchi, Lucas Mallada
ous studies carried out in the same area with other
5 (1993) 101.
biota samples (locust Orthopter orthopter) [7]. Also [5] C. Nerı́n, T. Polo, C. Domeño, I. Echarri, Int. J. Environ.
is important to consider the values obtained for Anal. Chem. 65 (1996) 83.
these compounds in the atmosphere and in the wa- [6] C. Nerı́n, T. Polo, J. Salafranca, J. Cacho, Int. J. Environ.
ter system in the same area [5,6]. Fig. 5 shows the Anal. Chem. 65 (1996) 27.
comparison. The bioaccumulation and bioconcentra- [7] C. Nerı́n, M.A. Cebrián, J. Cacho, Determinación de pesti-
cidas organoclorados en ortópteros pirenaicos, Incidencia
tion factors obtained are similar to those described sobre el Parque Nacional de Ordesa y Monte Perdido, in:
in the literature and suggests that the frogs could be Proceedings of X Reunión Nacional de la Sociedad Española
used as bioindicators of the contamination by these de Quı́mica Analı́tica, Almerı́a, 1997.
compounds. [8] J.L. Snyder, R.L. Grob, M.E. Mcnally, T.S. Oostdyk, Anal.
Chem. 64 (1992) 1940.
[9] A. Valverde Garcı́a, A.R. Fernández-Alba, Seminario Inter-
nacional sobre Residuos de Plaguicidas, Instituto de Estudios
6. Conclusions Almerienses, Universidad de Almerı́a, 1995.
[10] C. Nerı́n, R. Batlle, J. Cacho, Anal. Chem. 69 (1997) 3304.
SFE has been shown as a powerful technique for the [11] C. Nerı́n, R. Batlle, J. Cacho, J. Chromatogr. A 795 (1998)
extraction of organochlorine compounds in biota sam- 117.
[12] J.M. Carrasco, P. Cuñat, M. Martı́nez, R.M. Martı́nez, E.
ples, with clear advantages such as efficiency (higher Primo, Rev. Agroquim. Tecnol. Alim. 12 (1972) 583.
recoveries), time consuming, cost and environmen- [13] C. Porte, J. Albaiges, Arch. Environ. Contam. Toxicol. 26
tally friendly, versus Soxhlet procedures. However, (1993) 273.
it requires the optimization in depth since the ex- [14] S. Tanabe, K. Kannan, M.S. Tabucanon, C. Siriwong, Y.
traction behaviour is strongly affected by the type of Ambe, R. Tatsukawa, Environ. Pollut. 72 (1991) 191.
[15] K. Kannan, S. Tanabe, A. Ramesh, A. Subramanian, R.
sample.
Tatsukawa, J. Agric. Food Chem. 40 (1992) 518.
From the environmental point of view, the frog sam- [16] J.A. Garcı́a-Regueiro, I. Dı́az, J.M. Monfort, Rev. Agroquim.
ples under study showed the presence of organochlo- Tecnol. Alim. 27 (1987) 405.
rine pesticides and some metabolites, which confirms [17] E. González Pradas, A. Valverde Garcia, J.L. Martı́nez Vidal,
once again the atmospheric transport of pesticides ev- J.L. Garcı́a Sánchez, Ann. de Quı́m. 88 (1991) 486.
erywhere and their persistence in biota. On the other [18] S. Winter, B. Streit, Chemosphere 24 (1992) 1765.
[19] D.M. Holstege, D.L. Scharberg, E.R. Tor, L.C. Hart, F.D.
hand, frogs seem to be good candidates to be used as
Galey, J. AOAC Int. 77 (1994) 1263.
bioindicators of air pollution since the samples are not [20] C. Nerı́n, C. Domeño, I. Echarri, A. Tornes, Quı́m. Anal. 16
very difficult to handle and these animals are present (1997) 101.
in a wide variety of environments. [21] R.K. Juhler, Analyst 123 (1998) 1551–1556.
312 C. Nerı́n et al. / Analytica Chimica Acta 464 (2002) 303–312

[22] C. Nerı́n, A. Tornes, C. Domeño, J. Cacho, Fresenius J. Anal. [25] I. Echarri, C. Nerı́n, D.E. Wells, C. Domeño, J. Cacho,
Chem. 352 (1995) 364. Analyst 123 (1998) 421.
[23] G.J. Reimer, D. Noot, A. Suarez, Int. J. Environ. Anal. Chem. [26] D.E. Wells, I. Echarri, Anal. Chim. Acta 286 (1994) 431.
59 (1995) 91. [27] I.J. Barnabas, J.R. Dean, S.M. Hitchen, S.P. Owen, J. Chro-
[24] A. Valverde, A.R. Fernández, M. Contreras, A. Agüera, J. matogr. A 665 (1994) 307.
Agric. Food Chem. 44 (1996) 1780. [28] K. Ballschmiter, Pure Appl. Chem. 68 (1996) 1771.