Chronic oral exposure to field-realistic pesticide combinations via pollen and nectar: effects on feeding and thermal performance in a solitary bee

Pesticide use is one of the main causes of pollinator declines in agricultural ecosystems. Traditionally, most laboratory studies on bee ecotoxicology test acute exposure to single compounds. However, under field conditions, bees are often chronically exposed to a variety of chemicals, with potential synergistic effects. We studied the effects of field-realistic concentrations of three pesticides measured in pollen and nectar of commercial melon fields on the solitary bee Osmia bicornis L. We orally exposed females of this species throughout their life span to 8 treatments combining two neonicotinoid insecticides (acetamiprid, imidacloprid) and a triazole fungicide (myclobutanil) via pollen and sugar syrup. We measured pollen and syrup consumption, longevity, ovary maturation and thermogenesis. Although bees consumed larger amounts of syrup than pollen, pesticide intake via syrup and pollen were similar. At the tested concentrations, no synergistic effects emerged, and we found no effects on longevity and ovary maturation. However, all treatments containing imidacloprid resulted in suppressed syrup consumption and drastic decreases in thoracic temperature and bee activity. Our results have important implications for pesticide regulation. If we had measured only lethal effects we would have wrongly concluded that the pesticide combinations containing imidacloprid were safe to O. bicornis. The incorporation of tests specifically intended to detect sublethal effects in bee risk assessment schemes should be an urgent priority. In this way, the effects of pesticide exposure on the dynamics of bee populations in agroecosystems will be better assessed.

(n=3 per field), and then stored at -80°C. Hundreds of additional flowers were frozen at -80°C for later pollen collection. In October, these flowers were dried in an incubator at 37 °C for 24 h to facilitate pollen removal 1 . Anthers were then collected and the pollen was extracted using a sieve of 150μm pore size (melon pollen grain Ø = 50-100μm 2 ).
We used 140-250 flowers to obtain approximately 0.1g of pollen per sample (n=3 per field).

Multi-residue analyses of pesticides
The multi-residual analyses were carried out in the Laboratorio Químico Microbiológico de Sevilla an official accredited analytical testing company (www.lqmsa.com). High performance liquid chromatography with quantification and confirmation by triple-quadrupole mass spectrometer detector (HPLC-QQQ) and gas chromatography with triple-quadrupole mass spectrometer detector (GC-QQQ) were used to target more than 200 compounds. For the 3 pesticides used in our trials, imidacloprid and acetamiprid were analyzed by HPLC-QQQ and myclobutanil by GC-QQQ. The recoveries were over the detection limit (LOD) of 3ng/g for each analyte.

Pollen and nectar preparation
Pollen preparation was performed using a modified version of the QuEChERS methodology described by David et al. 3 . Pollen samples (100-200mg) were weighed and introduced in 15-ml Eppendorf Tubes with a ceramic microbead Two ml of water were added to each sample to form an emulsion and extracted by adding 2ml of acetonitrile (CH3CN, Baker 9012 "HPLC Analyzed), mixing in Agytax ® SR2 for 2 minutes and sonicating for 10 minutes. Then, 1g of magnesium sulphate/sodium acetate mix (4:1) (QuEChERS Salts, Agilent 5982-0650) was added followed by immediate shaking (Agytax ® SR2) for 2 min and sonication without heating for 10 min. The supernatant was transferred into a 4-ml polypropylene tube with a ceramic microbead and Agilent QuEChERS dispersive 5982-5056 was added to match the weight of the sample and shaken for 2 min (Agytax ® SR2) and centrifugated (3000 rpm for 5 min).
The supernatant was measured and transferred to a centrifuge tube and dried with high purity N 2 stream. The extract was reconstituted with 100μL of CH3CN.
Nectar preparation followed a similar procedure. To calculated nectar weight, nectar samples (ca. 50μL) were inserted in pre-weighed Eppendorf tubes. The volume of water added was adjusted to 1mL. Nitrogen was the nebulising gas with a flow rate of 9L·min -1 and temperature 345°C.

Instruments
Nebulizer gas pressure was set at 40 psi.
To perform the GC-QQQ, we used a GC Agilent Technologies GC7890A with an automatic liquid autosampler and a split-splitless injector. A DB-5 (5%Phenyl 95% Methylpolysiloxane) column Agilent 19091S-433 (30m × 0, 25mm × 0.25 µm) was used. A sample volume of 1 µL was injected into the GC in splitless mode at an injector temperature of 250°C. The oven temperature program was as follows: initial temperature 70 °C (held for 2 min) increased by 25°C/min to 150°C; increased by 3°C /min to 200°C (held for 1 min); increased by 8°C/min to 280°C (held for 10 min). High Purity Nitrogen gas was used as collision gas with a flow rate of 1.5mL/min and Helium at 1.35mL/min for quenching effect into the collision cell (octopole). The concentration of the calibration standards were 2 to 100µg/L