Novel indices reveal that pollinator exposure to pesticides varies across biological compartments and crop surroundings.

Declines in insect pollinators have been linked to a range of causative factors such as disease, loss of habitats, the quality and availability of food, and exposure to pesticides. Here, we analysed an extensive dataset generated from pesticide screening of foraging insects, pollen-nectar stores/beebread, pollen and ingested nectar across three species of bees collected at 128 European sites set in two types of crop. In this paper, we aimed to (i) derive a new index to summarise key aspects of complex pesticide exposure data and (ii) understand the links between pesticide exposures depicted by the different matrices, bee species and apple orchards versus oilseed rape crops. We found that summary indices were highly correlated with the number of pesticides detected in the related matrix but not with which pesticides were present. Matrices collected from apple orchards generally contained a higher number of pesticides (7.6 pesticides per site) than matrices from sites collected from oilseed rape crops (3.5 pesticides), with fungicides being highly represented in apple crops. A greater number of pesticides were found in pollen-nectar stores/beebread and pollen matrices compared with nectar and bee body matrices. Our results show that for a complete assessment of pollinator pesticide exposure, it is necessary to consider several different exposure routes and multiple species of bees across different agricultural systems.

An assessment of honeybee colony matrices, Apis mellifera (Hymenoptera: Apidae) to monitor pesticide presence in continental France.

The frequency of occurrence and relative concentration of 44 pesticides in apicultural (Apis mellifera) matrices collected from five French locations (24 apiaries) were assessed from 2002 to 2005. The number and nature of the pesticides investigated varied with the matrices examined-living honeybees, pollen loads, honey, and beeswax. Pollen loads and beeswax had the highest frequency of pesticide occurrence among the apiary matrices examined in the present study, whereas honey samples had the lowest. The imidacloprid group and the fipronil group were detected in sufficient amounts in all matrices to allow statistical comparisons. Some seasonal variation was shown when residues were identified in pollen loads. Given the results (highest frequency of presence) and practical aspects (easy to collect; matrix with no turnover, unlike with bees that are naturally renewed), pollen loads were the best matrix for assessing the presence of pesticide residues in the environment in our given conditions.

A survey of pesticide residues in pollen loads collected by honey bees in France.

In 2002, a field survey was initiated on French apiaries to monitor weakness of honey bee, Apis mellifera L., colonies. Apiaries were evenly distributed in five sites located on continental France. Five colonies were randomly selected in each apiary, leading to a total of 125 studied honey bee colonies. For 3 yr (starting in autumn 2002), colonies were visited four times per year: after winter, before summer, during summer, and before winter. Pollen loads from traps were collected at each visit. Multiresidue analyses were performed in pollen to search residues of 36 different molecules. Specific analyses were conducted to search fipronil and metabolites and also imidacloprid and metabolites. Residues of 19 searched compounds were found in samples. Contamination by pesticides ranged from 50 to 0%. Coumaphos and tau-fluvalinate residues were the most concentrated of all residues (mean concentrations were 925.0 and 487.2 microg/kg, respectively). Fipronil and metabolite contents were superior to the limit of detection in 16 samples. Residues of fipronil were found in 10 samples. Nine samples contained the sulfone compound, and three samples contained the desulfinyl compound. Residues of imidacloprid and 6-chloronicotinic acid were found in 69% of samples. Imidacloprid contents were quantified in 11 samples with values ranging from 1.1 to 5.7 microg/kg. 6-Chloronicotinic acid content was superior to the limit of quantification in 28 samples with values ranging from 0.6 to 9.3 microg/kg. Statistical tests showed no difference between places of sampling with the exception of fipronil. Possible origins of these contaminations, concentration and toxicity of pesticides, and the possible consequences for bees are discussed.