Introduction

Farmers around the world use pesticides as an insurance policy against the possibility of a devastating crop loss from pests and diseases. The steady growth of the world's population necessitates increased food production, which in its turn places the onus upon the international chemical industry to find new compounds to protect crops against pests and other biological threats (Cieslik et al. 2011).

According to data from the EU's Pesticide Action Network, as of 2008, some 350 different pesticides were detected in food produced in the EU. More than 5 % of products contained pesticides at levels exceeding the EU's maximum permitted level (Fenik et al. 2011).

The presence of these compounds in food could constitute a serious risk to human and animal health and environment. Different organizations have established strict regulation controls on the pesticide handling to minimize the exposure of the population. Maximum residue levels (MRLs) are set by the European Commission to protect consumers from exposure to unacceptable levels of pesticide residues in food and feed (Castillo et al. 2011). Integrated pest management (IPM) cultivation includes guidelines that farmers use to enforce actions for the production of safe agricultural products. Organic farming (OF) distinguishes from all other forms of farming by a rejection of soluble minerals as fertilizers and synthetic pesticides in favor of natural ones. IPM and OF cultivation have a higher efficiency in terms of product safety and quality. New research has been in progress and shown that organic farming actually yields better results than other techniques. A study with model estimates indicates that organic methods could produce enough food on a global per capita basis to sustain the current human population (Badgley et al. 2007).

To advise IPM and conventional farmers, the Portuguese department of agriculture published a pesticide list allowed in the production of several crops, including strawberries (Lopes and Simões 2006). Over the past few years, the European organic food market has experienced annual growth rates of more than 10 % (Stolz et al. 2011).

The new regulatory frameworks require sensitive and highly specific methods for the measurement of multiclass pesticide residues (Lehotay et al. 2011). A number of sample preparation techniques, and methods of analyses, have been developed for the pesticide residue determination in wide range of foodstuffs and other agricultural products (Fernandes et al. 2011a). Traditionally, pesticide residue analyses are carried out in a sequence of several steps: extraction by organic solvent methods such as liquid–liquid partitioning (Li et al. 2008; Zamora et al. 2004), solid-phase microextraction (Beltran et al. 2003; Cortes-Aguado et al. 2008; Lambropoulou and Albanis 2003), matrix solid-phase dispersion (Albero et al. 2003; Barriada-Pereira et al. 2010; Radisic et al. 2009), supercritical fluid extraction (Rissato et al. 2005a, b; Saito et al. 2004), normal phase liquid chromatography and/or gel permeation chromatography (Cajka et al. 2008, Hoh et al. 2009), solid-phase extraction (Albero et al. 2005; Barriada-Pereira et al. 2010; Stajnbaher and Zupancic-Kralj 2003), and dispersive solid-phase extraction (d-SPE) (Anastassiades et al. 2003; Castillo et al. 2011; Lehotay 2011; Zhang et al. 2011).

An alternative technique is Quick, Easy, Cheap, Effective, Rugged, and Safe (QuEChERS) developed by Anastassiades et al. (2003). QuEChERS is a sample preparation approach entailing solvent extraction of high-moisture samples with acetonitrile, ethyl acetate, or acetone and partitioning with magnesium sulfate alone or in combination with other salts followed by cleanup using d-SPE. Using QuEChERS, analytical chemists can prepare about ten samples in a 30-min period, with a low cost. This can decrease material costs and less than 10 mL of organic solvent is used with QuEChERS/sample (Cieslik et al. 2011). Two versions of the QuEChERS method have been successfully validated in multiple laboratories through the auspices of AOAC International (Lehotay 2007) and European Committee for Standardization (Version p 2007-10-24). There are also many modified versions of the QuEChERS approach with small advantages and disadvantages mainly depending on personal preferences (Lehotay et al. 2011).

Multiresidue approaches, with low quantification limits, are a rapid methodological answer. Chromatographic analysis traditionally used gas chromatography (GC) with electron capture detection (Fernandes et al. 2011b; Furlani et al. 2011) and nitrogen–phosphorus detection (Aysal et al. 2007; Barakat et al. 2007) for pesticides. Nowadays, it is preferable to use mass spectrometry (MS) (Anastassiades et al. 2003) or tandem mass spectrometry (MS/MS) detection in order to verify peak identity. GC-MS/MS offers various advantages in selectivity and sensitivity at low concentrations in the most difficult matrixes (Meng et al. 2010). Automated large volume injection systems, based on programmable temperature vaporizing injection, are employed to improve the limits of detection for pesticides in food matrices (Banerjee et al. 2010; Cajka et al. 2008).

Owing to the diverse use of strawberries and increasing awareness towards OF, it becomes important to evaluate the organically produced crops. A comparison of varieties grown under organic and IPM conditions for pesticides evaluation is therefore a good approach.

The purposes of this study were: (1) to evaluate a selective, sensitive, quick, and easy sample treatment, based on QuEChERS procedure for strawberries analysis; (2) to evaluate the efficiency of d-SPE cleanup step and to design an instrumental multiresidue method for the quantitative analysis at trace level of 25 pesticides (included in the published list from Portuguese Government for strawberry crops and suitable for GC-MS analysis) using GC-MS/MS; and (3) to monitor multiclass pesticide residues in strawberries originating from IPM and OF in order to control the efficiency of these specific cultivation system, for safety purposes.

Materials and methods

Reagents

A total of 25 pesticides as listed in Table 1 were included in the analytical method. All pesticide standards had purity of ≥95 % (typically > 99 %) and were obtained from Sigma-Aldrich, Riedel-de Häen, and Chem Service. Hexane and methanol were chromatographic grade and were supplied by Merck (Darmstadt, Germany). Stock standard solutions (around 1,000 μg/L) were prepared by dissolving reference standards in n-hexane and methanol and were stored in a freezer at 4 °C. Working pesticide standard mixtures were prepared by dilution of stock solutions in hexane. The 4,4′-dichlorobenzophenone was used as an internal standard (IS).

Table 1 GC-MS/MS conditions and limits of the method (LOD and LOQ)
Full size table

QuEChERS material was obtained from UCT (Bristol, PA, USA). The selected QuEChERS was the citrate version with 6 g anhydrous magnesium sulfate, 1.5 g sodium chloride, 1.5 g trisodium citrate dehydrate, and 0.75 g disodium hydrogenocitrate sesquihydrate (Ref. EUMIV50CT-VP, UCT) and a cleanup with 150 mg primary secondary amine (PSA), 150 mg MgSO4, and 50 mg C18 (Ref. CUMPS15C18CT, UCT).

Strawberry samples

OF and IPM strawberries were collected in the first week of May in two consecutive years (2009 and 2010) from crops in the center of Portugal. Different varieties of strawberries were collected including Siba, Camarosa, Festival, and Albion in both types of farming. The samples were analyzed following the procedure described below. The samples with no presence of target analytes were used as blank strawberry samples (variety Camarosa from OF) in the preparation of standards and in the recovery studies.

GC-MS/MS analysis

In this study, the GC-MS/MS analysis was performed on a TRACE GC Ultragas chromatograph Polaris Q coupled with ion trap mass spectrometer (Thermo Fisher Scientific) operated in the electron impact ionization at 70 eV controlled by Xcalibur 1.3 software. Injection was conducted by an autosampler (AI3000) in combination with a split/splitless mode. Helium (Linde Sógas purity ≥99.999 %) was used as a carrier gas at 1 mL/min and the injection volume was 2 μL.

The analytes were separated on a ZB-XLB capillary column from Phenomenex® (30 m × 0.25 mm × 0.25 μm). The column oven temperature was programmed as follows: initial temperature was 40 °C (held for 1 min), increased by 30 °C/min to 220 °C (held for 5 min), increased by 10 °C/min to 250 °C and held at this temperature for 20 min and finally increased again by 5 °C/min to 285 °C and held at this temperature for 5 min. The ion source temperature was 250 °C and the electron multiplier at 1,900 V (autotune to gain 1 × 107). The pesticides involved in this study were identified by comparing the retention time and three ions (one target and two qualifiers) with the MS2-NIST library.

The MS/MS conditions were fixed for each compound, trying to select as precursor ion the one with the highest m/z ratio and abundance (Table 1). The limit of detection (LOD) was calculated as three times higher than the level of noise (1), and the limit of quantification (LOQ) was defined to ten times of the noise level (2).

$$ {\text{LOD}} = {3}C/\left( {S/N} \right) $$
(1)
$$ {\text{LOQ}} = {1}0C/\left( {S/N} \right) $$
(2)

where S/N is the signal-to-noise and C is the lowest spiking level concentration.

The one-way analysis of variance (ANOVA) was used in data analysis to estimate the influence of the residual standard deviation (RSD) between OF and IPM strawberry samples. The differences between groups were considered significant when p < 0.05.

Sample preparation

For the initial extraction step, an amount (10 g) of chopped strawberries was weighted into a 50-mL centrifuge tube and 50 μL of IS solution was added. The strawberry sample was left for 30 min at room temperature to let the n-hexane evaporate before the addition of 10 mL of acetonitrile.

The resulting solution was shaken for 1 min followed by the addition of citrate version of QuEChERS (EUMIV50CT-VP). The centrifuge tube was capped and shaken vigorously for 1 min to prevent salt agglomeration before centrifugation at 3,000 rpm for 5 min at room temperature. An aliquot of 1.5 mL was sampled from the upper layer into a 2-mL cleanup vial (CUMPS15C18CT) and again vortexed for 1 min and then centrifuged for 5 min at 4,000 rpm at room temperature. From the upper layer of the prepared sample, an aliquot of 1.0 mL was transferred into a vial and put under a mild stream of nitrogen to dryness. Finally, 1 mL of hexane was added to dissolve the residue and then 2 μL of this solution was injected onto the gas chromatograph. Six matrix standards were used for matrix-matched calibration standards that included all 25 pesticide analytes at 20, 50, 100, 200, 400, and 500 μg/kg equivalents.

For recovery studies, 10 g of strawberry sample free of detectable pesticides was spiked and homogenized at 50, 200, and 400 μg/kg levels for each pesticide. The mixture was left for 30 min before the beginning of the extraction process. Samples were then prepared according to the procedure aforementioned.

Results and discussion

Chromatographic analysis

The proposed method attempts to evaluate pesticide residues in strawberry extracts from OF and IPM using the advantages of MS/MS detection. The GC-MS/MS conditions including precursor ion and qualifier ions (Q) of all target compounds and retention time (tR) are shown in the Table 1.

The chromatogram (Fig. 1) shows a separation of the 25 selected pesticides, with tR in the range of 7–38 min (Table 1). Considering that strawberry samples are complex, the matrix effects are expected which may have a significant impact in the results of pesticide quantification. For this study, the matrix effects were examined by comparing spiked matrix-matched standards to solvent-based standards. Calibration curves were constructed by plotting the peak area against the analyte standard concentrations. A calibration curve was constructed for each compound using six different concentrations of 20, 50, 100, 200, 400, and 500 μg/kg. No evidence for nonlinearity was observed for all pesticides in the range of concentrations selected for strawberry matrix-matched calibration, as all R 2 values were higher than 0.99. The solvent-based and matrix-matched calibration curves were compared for all compounds. Both calibration curves were linear and the determination coefficients were always higher than 0.99 for all pesticides. However, the effect of the matrix was observed for most of the tested pesticides. Figure 2 shows the matrix suppression result for an example: malathion. As a conclusion, the slope of calibration curve changes from 130 to 219 for direct calibration and matrix-matched standards, respectively. The matrix-matched calibration should be employed rather than straightforward solvent calibration. Consequently, matrix-matched standard calibrations were used along this work.

Fig. 1
figure 1

GC-MS/MS chromatograms of the Camarosa OF strawberry extract (400 μg/kg) of each pesticide that was cleaned up with dispersive solid-phase extraction. Peak identification in order of increasing tR: 1—methiocarb; 2 and 3—diazinon and dazomet; 4—pyrimethanil; 5—vinclozolin; 6—malathion; 7—chlorpyrifos; 8—tetraconazole; 9 and 10—cyprodinil and pendimetaline; 11 and 12—tolylfluanid and procymidone; 13 and 14—captan and folpet; 15—mepanipyrim; 16 and 17—fludioxonil and bupirimate; 18—fluazifop-p-butyl; 19—myclobutanil; 20 and 21—bifenthrin and fenhexamid; 22—iprodione; 23—quizalofop-p-ethyl; 24—deltamethrin; 25—azoxystrobin

Full size image
Fig. 2
figure 2

Experimental matrix effect for malathion (matrix suppression) in Camarosa OF strawberry sample

Full size image

LOD and LOQ were calculated from the S/N obtained by analyzing spiked strawberry sample in 50 μg/kg level. Seven compounds had a LOD lower than 1 μg/kg, whereas 12 had a LOQ below 5 μg/kg. Although the obtained LOD and LOQ for most of the pesticides met the 10-μg/kg threshold requirement for OF foodstuffs, the LOQ of five pesticides (captan, dazomet, fludioxinil, folpet, and quizalofop-p-ethyl) did not meet this requirement (Table 1). Concerning the MRL established by the European Commission (SANCO/12495/2011) for each pesticide, the LOQ obtained is sufficiently low for the developed method, which allows its use for monitoring purposes.

QuEChERS and cleanup

The QuEChERS method is well described for food and vegetable samples. However, some preliminary tests were made and the optimal conditions described by the EN 15662 method were chosen.

The acetonitrile extracts were heavily pigmented, containing large amounts of endogenous interfering compounds that were coextracted with the pesticides. Furthermore, the sample matrix coextractants may have a deleterious effect on the capillary columns, may interfere with the detection of pesticides present at trace levels, and/or may result in a sample matrix-induced enhancement effect. Therefore, QuEChERS was necessary for further cleanup prior to chromatographic analysis. The cleanup composed with MgSO4, C18, and PSA was assessed based on the efficiency of cleanup and recovery evaluation. The extract obtained after the cleanup step was translucent because the anthocyanin color was retained in PSA sorbent.

No differences were observed in extraction efficiencies using IPM and OF samples. Table 2 shows the average recoveries of the 25 pesticides fortified at 50, 200, and 400 μg/kg. In both strawberry samples (OF and IPM), the recoveries were similar at the levels of 200 and 400 μg/kg (from 70 to 154 %) and 50 μg/kg (from 60 to 110 %) (Fig. 3). In the case of the lower spiking level (50 μg/kg), some of the pesticides showed a decrease in recovery, instead at higher levels. In 200 and 400 μg/kg spiking levels, the signal was enhanced in some analytes. The majority of the pesticides (70–119 %) gave satisfactory recoveries (ranging from 70 to 120 %); diazinon and tolylfluanid gave a recovery lower than 70 % in the spiking level of 50 μg/kg in IPM and OF strawberries. Fludioxinil, mepanipyrim, myclobutanil, and tetraconazole showed recoveries higher than 120 % in the spiking level of 200 and 400 μg/kg. Despite the diazinon and tolylfluanid in the lower spiking level, and mepanipyrim, myclobutanil, tetraconazole, and fludioxonil in the higher spiking level, the others had satisfactory recoveries (in a range from 70 to 120 %) after cleanup. In the spiking level with 200 μg/kg, all the pesticides are in the range 70–120 % defined by SANCO (SANCO/12495/2011).

Table 2 Recoveries obtained in three different spiking levels (50, 200, and 400 μg/kg) from OF and IPM Camarosa strawberries (n = 3)
Full size table
Fig. 3
figure 3

Comparison of recovery values for IPM and OF Camarosa strawberry samples. Results obtained for studies conducted after fortification at the level of 50 μg/kg

Full size image

The repeatability, expressed as the RSD of the spiked sample concentrations, was significantly better, but more varied, for the IPM strawberries, about 1 to 7 % than for the OF strawberries, at about 2 to 12 %. At the lower concentration level, the repeatability for OF obtained data ranging from 4 to 12 % (OF) instead of 1 to 6 % in the IPM samples. Descriptive statistics and one-way ANOVA were used in the data analysis of the RSD (in percent) to evaluate differences between the RSDs from OF and IPM. The RSDs from OF were higher than the RSDs from IPM because the F value (F = 67) from ANOVA test is higher than the critical F (F critical = 4) showing that the RSDs between the two types of agriculture are statistically different. These data suggest a larger variability in samples from OF. The authors propose that the less controlled crop (OF) leads to more crop variability and hence more matrix variability. The RSDs (<12 %) were obtained for all samples (n = 3) at 50, 200, and 400 μg/kg (Table 2).

Strawberry samples

Strawberry samples collected from various sources were prepared in replicates of three and analyzed by the method validated in this work. In total, 12 samples from which nine matrices from OF and three from IPM were prepared and analyzed. Various pesticides in the three IPM samples at various concentrations are shown in Fig. 4. The results showed the presence of fludioxonil, bifenthrin, mepanipyrim, tolylfluanid, cyprodinil, tetraconazole, and malathion below the MRL in 2009 strawberries grown using IPM. In 2010, only iprodione (MRL = 15,000 μg/kg) and cyprodinil (MRL = 5,000 μg/kg) were detected in IPM strawberry samples. The highest concentration was obtained for iprodione (1071 μg/kg) in a 2010 IPM strawberry sample. The determined pesticide (iprodione) is in agreement with previous findings related with strawberry (Húsková et al. 2009). None of the pesticide found would be considered a violation of permitted uses. No pesticides were found in organic farming strawberry samples.

Fig. 4
figure 4

Results from IPM strawberry samples from 2009 and 2010 crops. MRL in Europe for strawberries: fludioxonil 3,000 μg/kg; iprodione 15,000 μg/kg; bifenthrin 500 μg/kg; mepanipyrim and its metabolite (2-anilino-4-(2-hydroxypropyl)-6-methylpyrimidine) 2,000 μg/kg; tolylfluanid (sum of tolylfluanid and dimethylaminosulfotoluidide expressed as tolylfluanid) 5,000 μg/kg; cyprodinil 5,000 μg/kg; tetraconazole 200 μg/kg; malathion (sum of malathion and malaoxon expressed as malathion) 20 μg/kg

Full size image

Conclusions

A simple and rapid method was developed to determine 25 pesticide residues allowed by law, if necessary, on strawberry samples. This method using QuEChERS sample preparation and GC-MS/MS analysis showed a high sensitivity and confirmatory power. To compensate for the matrix-induced response enhancement, the use of fortified blank samples as calibration standards was introduced.

The final method has good linearity and an LOD of 0.1 μg/kg for bifenthrin and tolylfluanid or higher was obtained by the others (lower than the MRLs established by the EU). The majority of the pesticides had recoveries ranging from 70 to 120 % at concentrations of 20–500 μg/kg. In the present study, 12 different Portuguese strawberry samples produced by OF and IPM were compared and no pesticides were found in OF strawberry samples. The proposed method is acceptable for monitoring purposes and allows the simultaneous analyses of a large number of pesticides with good recoveries, low detection limits, and high throughput.