2.1. Study Design Consideration
The field experiments were designed to meet the objective of the study described above, as well as to comply with the respective Organization for Economic Co-operation and Development (OECD) guidelines for studies of occupational exposure to pesticides [
18].
Six field trials (identified as trials I–VI) were conducted in three greenhouses (GHs) to determine operator exposure. In each trial, one M/L and one application task was considered, both carried out by the same operator. It is noted that the term “operator” is used in the present work to define the volunteers participating in the study, regardless of their M/L or application tasks. Thus, to specifically address operators during their M/L or application tasks, the terms “mixer/loader” and “applicator” are used, respectively, throughout the text hereafter.
In total, three operators took part in the study, each of whom carried out a set of two trials, i.e., one in the morning and one in the afternoon. To isolate the trials, each time, one operator treated one GH. More specifically, the first three trials (identified as trials I, II and III) involved monitoring of M/L and application of one pesticide (pesticide A) in the morning. Respectively, trials IV, V, VI involved monitoring of M/L and application of a different pesticide (pesticide B) in the afternoon; therefore, corresponding to re-entry treatment. In all six trials conducted, the exposure was determined separately for mixer/loaders and applicators.
The study design and experimental scheme followed are presented in
Table 1.
2.2. Field Phase
All field trials took place on the island of Crete, Greece (Tympaki region). Greenhouse tomato crops were selected for the applications, grown with standard protected cropping systems for the region, e.g., narrow inter-row width. The treated area in each one of the three GHs was 0.13 ha, while the duration of application ranged between 36 to 47 min (see data in
Table 2). The spray liquids (s.l.) were applied using a spray gun connected via a hose to the tank. The distance of the operators from the tank was sufficient (>2 m) to avoid any cross-contamination due to the proximity to the tank. Field trial parameters and data are presented in
Table 2. In addition, since applications took place indoors, the impact of environmental conditions was negligible.
More specifically, the procedure implemented was based on the principles of the whole-body dosimetry (WBD) method [
8,
18,
19,
20], as adapted and designated in detail in a previous study [
11]. The coverall type used was a cotton coverall (100% cotton, see
Figure 1), serving both as operator’s standard working clothing and concurrently as an outer exposure dosimeter. Inner dosimeters (100% cotton shirts and long pants) were used for the monitoring of actual exposure. Head and hand exposure were measured in accordance with the sampling method already used in a previous study [
10] using cotton caps and gloves (inner cotton, outer nitrile) as dosimeters, respectively. Quality control samples of all dosimeters used were fortified in the field at two fortification rates (plus blanks) as a measure of the active substance (a.s.) stability and recovery according to the previously described procedure [
10,
21]. The nitrile gloves extraction, both for field- and spiked samples, was conducted in the field immediately after sampling, as this matrix is known to retain analytes, impacting recovery efficiency if stored for any length of time [
22]. All samples were stored in freezers (−18 °C) shortly after the trial. Exposure term definitions (potential, actual, etc.) and clarifications regarding the various dosimeters considered for the respective exposure calculations are provided in
Section 2.4.
Two different pesticides (authorized and normally used either as a tank mix or in sequential application on the same day) were applied. Mixing and loading (M/L) and application procedures were in accordance with local practices and the principles of Good Agricultural Practice (GAP). The experimental work followed the design and sampling scheme described in the respective section above and depicted in
Table 1. The pesticide A was an emulsifiable concentrate (EC) fungicide formulation containing the a.s. bupirimate at 25%
w/
v. Since pesticide A was the first one applied, it also served as the pesticide tested for the study of the potential transfer of residues from the equipment and the treated crop to the operator. Pesticide B was a suspension concentrate (SC) fungicide formulation containing the a.s. tebufenozide at 24%
w/
v. Both pesticides were purchased from a local distributor and the respective containers were checked for expiry dates and intact packaging by the field scientists, and found to comply with the quality requirements and foreseen label.
During the application of the first pesticide (pesticide A), in the morning, the potential dermal exposure (PDE) and the actual dermal exposure (ADE) for the M/L and for the applicators were measured using standard whole-body techniques (see
Section 2.4 below) following published procedures [
11]. Τhe volunteers wore outer coveralls (jacket and trousers 100% cotton) and inner coveralls (long-sleeved T-shirt and long johns, 100% cotton).
In the course of application of the second pesticide (pesticide B), the afternoon of the same day, a new clean set of dosimeters was provided to the operators, and the dermal exposure to pesticide B, but also to pesticide A, applied in the morning, due to potential transfer of residues from the equipment and the already treated crop, was measured. The re-entry of the operators to apply pesticide B took place six hours after the application of pesticide A, which was adequate time for drying of the spray solution (also confirmed by visual inspection of the crop).
2.3. Laboratory Phase
The laboratory phase of the study was carried out at the Laboratory of Pesticides’ Toxicology of Benaki Phytopathological Institute, Greece. Stock and working solutions (used for the validation of analytical method) of the analytical standards tebufenozide (obtained from ChemService 99.5% purity) and bupirimate a.s. (acquired from ChemService, 99.5% purity) were prepared as described in previous related publication [
12]. All the aforementioned solutions were stored at −18 °C.
For both pesticides used, the a.s. residues from the dosimeters were extracted with methanol according to the procedure described earlier [
10]. Especially for the outer gloves (nitrile), the extraction had already been performed in the field after the end of each application based on the procedure defined in the aforesaid study, since it was known that recovery of the a.s. reduces over time with this matrix [
23]. The chemical analysis (liquid chromatography electrospray mass spectrometry, LC-ESI/MS) was conducted using previously described conditions and the same instrumentation [
12].
Considering that no matrix effect from fabric and glove dosimeters was observed during the validation process (for this purpose, a series of calibration standard solutions was fortified with 40% concentrated extract of cotton matrix and analyzed, showing no evidence of influence on the detector response), all the standard solutions prepared and used in the analyses were prepared in methanol. The calibration standards were mixtures of equivalent concentration of both a.s. in this solvent. As the analytical routine procedure for completion of all the sample extraction and measurements took several weeks, calibration curves were prepared every two weeks, during this period, to ensure that possible baseline drifting and instrumental variations were detected. Acceptable linearity (r2 > 0.997) was observed in the range of 0.001 μg/mL to 0.050 μg/mL for bupirimate and up to 1 μg/mL for tebufenozide. The rest of validation criteria (accuracy, precision, specificity) were fulfilled as well.
The limit of quantification (LOQ) values for fabric (inner, outer) and glove dosimeters were equal to the lowest amount of a.s. that could be extracted and analyzed from a given dosimeter surface area with satisfactory recovery (>70%), providing a final extract solution concentration not less than the lowest point of the calibration curve (i.e., 0.001 μg/mL). Thus, LOQs corresponded for both a.s. to 0.007 μg/cm2 for coverall fabrics and 0.6 μg/dosimeter for cap and for nitrile glove.
2.5. Quantitative Risk Assessment
To proceed to a quantitative risk assessment, the systemic exposure (
SE) in mg kg
−1 bw day
−1 was calculated using Equation (1), i.e., based on the measured ADE and considering a dermal penetration factor (
Pf) for the a.s.
ADE = mass (mg) of the a.s. per kg of a.s. applied;
AR = application rate, kg a.s./ha;
TA = treated area, i.e., 1 ha; in line with EFSA Guidance of 2022 (see below);
Pf = penetration factor (%);
BW = body weight (kg), 60 kg.
Then, the hazard quotient (
HQ) was calculated (Equation (2)) by dividing
SE with a health-based guidance value (
HBGV, in this case the acceptable operator exposure level (AOEL)).
When HQ is greater than 1, then an unacceptable risk is concluded.
In case of combined exposure to more than one pesticide, the hazard index (HI) was calculated by adding the individual HQs. Again, when HI is greater than 1, then an unacceptable risk is concluded for the combined exposure.