| dc.description.abstract | Pesticide use in agriculture is one of the main factors driving pollinator declines. Bees are highly important pollinators and may be at risk from exposure to pesticide compounds while foraging for pollen and nectar. In this thesis, I explored the extent of pollen and nectar contamination by pesticides used in agriculture.
Worldwide knowledge gaps were identified via a novel systematic literature review, which summarised and critically evaluated pesticide residue levels and occurrence in nectar and pollen directly collected from plants. Through this review, it became clear that most previous work had focussed on detecting neonicotinoid insecticide residues, more often examining pollen rather than nectar, and that residues of several compounds were estimated to exceed the lethal dose for several bee species. However, studies were geographically constrained, and the range of compounds evaluated so far was narrow.
Given this insufficient understanding of the extent to which different compounds may be detected in pollen and nectar, I compared the mixtures and levels of pesticide residues in pollen and nectar of crops and wild plants growing in the margins of the crops in Ireland, whose land use is dominated by agriculture. Based on their application (volume and frequency) and relevance to bee health, the fungicides azoxystrobin, boscalid and prothioconazole, the herbicides fluroxypyr and glyphosate (and its main metabolite AMPA), and five neonicotinoid insecticides (acetamiprid, clothianidin, imidacloprid, thiacloprid and thiamethoxam), were identified as the targeted analytes for this residue evaluation study. As model crop species I used oilseed rape (Brassica napus L.), while blackberry plants (Rubus fruticosus agg.) were used as an example wild plant species.
Overall, more residues were detected in B. napus pollen and nectar than in the wild plant, and B. napus pollen had the highest mean concentration of residues. My results demonstrated that most detections were from fields with no recorded application of the respective compounds in that year, but higher concentrations were observed in recently treated fields. The most frequently detected compounds were fungicides, and all matrices were contaminated with at least three compounds. The most common compound mixtures were two fungicides and one neonicotinoid insecticide not recently applied on the cultivated fields. This indicated that persistent compounds like the neonicotinoids, should be continuously monitored for their presence and fate in the field environment.
Following this, I compared the mixtures and levels of pesticide residues in pollen of different crops and corbicular pollen from honey bees and bumble bees foraging simultaneously under the same agricultural conditions. For this, as model crop species I used oilseed rape (Brassica napus L.) and broad bean (Vicia faba L.), while pollen foragers of the common honey bee (Apis mellifera L.) and various wild bumble bee species (e.g., Bombus lucorum agg., B. lapidarius etc.) were collected from crops.
Samples from B. napus fields had higher number of pesticide detections, higher residue concentrations and higher number of different pesticide compounds detected compared to those from V. faba fields and the compounds azoxystrobin, boscalid and thiamethoxam formed the most common pesticide mixture. Crop pollen was contaminated only with fungicides, honey bees collected pollen mainly with fungicides, while bumble bee pollen had more insecticide detections. The highest number of compounds and most detections were observed in bumble bee collected pollen, where notably, all the five neonicotinoids assessed (acetamiprid, clothianidin, imidacloprid, thiacloprid, and thiamethoxam) were detected despite the European ban and no recent application on the studied fields. The concentrations of neonicotinoid insecticides were positively correlated with the number of wild plant species present in the bumble bee pollen samples. These results raise concerns about potential bee exposure to multiple residues and whether honey bees are suitable surrogates for pesticide risk assessments for all bee species.
Finally, since glyphosate and its main metabolite AMPA have different physicochemical properties compared to the rest of the evaluated compounds, I evaluated them using separate analytical methods specifically designed for polar pesticides that I optimised. Their occurrence and levels in crop and wild plant pollen and nectar, as well as honey bee and bumble bee collected pollen under realistic agricultural conditions were evaluated for the first time. My results indicate off-target contamination of non-target plant pollen and nectar raising concerns for possible implications of glyphosate residues detected in nectar, connected to bee and human health.
This work establishes a baseline of knowledge on the pesticide contamination levels of matrices relevant to bee species, such as pollen and nectar directly collected from plants, and honey bee and bumble bee corbicular pollen. Future research can draw upon the findings of this research and expand our understanding of pesticide residue fate in an agricultural context and assess whether bees and other beneficial insects are affected by pesticide exposure, and to which extent. This consequently can apprise ecologically friendly agricultural practices and inform policies, which will contribute to a sustainable pesticide use for future generations. | en |
