Metals in soil are known to negatively affect the health of many groups of organisms, but it is unclear whether they can affect plant-pollinator interactions, and whether pollinators that visit plants growing on contaminated soils are at risk of ingesting potentially toxic resources. We address whether the presence of metals in nectar alters foraging behavior by bumblebees by manipulating nectar with one of two common soil contaminants (Al or Ni) in flowers of Impatiens capensis (Balsaminaceae). While the presence of Al in nectar did not influence foraging patterns by bumblebees, flowers containing Ni nectar solutions were visited for shorter time periods relative to controls, and discouraged bees from visiting nearby Ni-contaminated flowers. However, because bumblebees still visited these flowers, they likely ingested a potentially toxic resource. Our findings suggest that soil metals could cascade to negatively affect pollinators in metal contaminated environments.

The increased use of pesticides has caused concern over the possible direct association of exposure to combinations of these compounds with bee health problems. There is growing proof that bees are regularly exposed to mixtures of agrochemicals, but most research has been focused on managed bees living in farmland, whereas little is known about exposure of wild bees, both in farmland and urban habitats. To determine exposure of wild bumblebees to pesticides in agricultural and urban environments through the season, specimens of five different species were collected from farms and ornamental urban gardens in three sampling periods. Five neonicotinoid insecticides, thirteen fungicides and a pesticide synergist were analysed in each of the specimens collected. In total, 61% of the 150 individuals tested had detectable levels of at least one of the compounds, with boscalid being the most frequently detected (35%), followed by tebuconazole (27%), spiroxamine (19%), carbendazim (11%), epoxiconazole (8%), imidacloprid (7%), metconazole (7%) and thiamethoxam (6%). Quantifiable concentrations ranged from 0.17 to 54.4 ng/g (bee body weight) for individual pesticides. From all the bees where pesticides were detected, the majority (71%) had more than one compound, with a maximum of seven pesticides detected in one specimen. Concentrations and detection frequencies were higher in bees collected from farmland compared to urban sites, and pesticide concentrations decreased through the season. Overall, our results show that wild bumblebees are exposed to multiple pesticides when foraging in agricultural and urban landscapes. Such mixtures are detected in bee tissues not just during the crop flowering period, but also later in the season. Therefore, contact with these combinations of active compounds might be more prolonged in time and widespread in the environment than previously assumed. These findings may help to direct future research and pesticide regulation strategies to promote the conservation of wild bee populations.

Neonicotinoid pesticides were first introduced in the mid-1990s, and since then, their use has grown rapidly. They are now the most widely used class of insecticides in the world, with the majority of applications coming from seed dressings. Neonicotinoids are water-soluble, and so can be taken up by a developing plant and can be found inside vascular tissues and foliage, providing protection against herbivorous insects. However, only approximately 5% of the neonicotinoid active ingredient is taken up by crop plants and most instead disperses into the wider environment. Since the mid-2000s, several studies raised concerns that neonicotinoids may be having a negative effect on non-target organisms, in particular on honeybees and bumblebees. In response to these studies, the European Food Safety Authority (EFSA) was commissioned to produce risk assessments for the use of clothianidin, imidacloprid and thiamethoxam and their impact on bees. These risk assessments concluded that the use of these compounds on certain flowering crops poses a high risk to bees. On the basis of these findings, the European Union adopted a partial ban on these substances in May 2013. The purpose of the present paper is to collate and summarise scientific evidence published since 2013 that investigates the impact of neonicotinoids on non-target organisms. Whilst much of the recent work has focused on the impact of neonicotinoids on bees, a growing body of evidence demonstrates that persistent, low levels of neonicotinoids can have negative impacts on a wide range of free-living organisms.

We assessed the state of knowledge regarding the effects of large-scale pollution with neonicotinoid insecticides and fipronil on non-target invertebrate species of terrestrial, freshwater and marine environments. A large section of the assessment is dedicated to the state of knowledge on sublethal effects on honeybees (Apis mellifera) because this important pollinator is the most studied non-target invertebrate species. Lepidoptera (butterflies and moths), Lumbricidae (earthworms), Apoidae sensu lato (bumblebees, solitary bees) and the section “other invertebrates” review available studies on the other terrestrial species. The sections on freshwater and marine species are rather short as little is known so far about the impact of neonicotinoid insecticides and fipronil on the diverse invertebrate fauna of these widely exposed habitats. For terrestrial and aquatic invertebrate species, the known effects of neonicotinoid pesticides and fipronil are described ranging from organismal toxicology and behavioural effects to population-level effects. For earthworms, freshwater and marine species, the relation of findings to regulatory risk assessment is described. Neonicotinoid insecticides exhibit very high toxicity to a wide range of invertebrates, particularly insects, and field-realistic exposure is likely to result in both lethal and a broad range of important sublethal impacts. There is a major knowledge gap regarding impacts on the grand majority of invertebrates, many of which perform essential roles enabling healthy ecosystem functioning. The data on the few non-target species on which field tests have been performed are limited by major flaws in the outdated test protocols. Despite large knowledge gaps and uncertainties, enough knowledge exists to conclude that existing levels of pollution with neonicotinoids and fipronil resulting from presently authorized uses frequently exceed the lowest observed adverse effect concentrations and are thus likely to have large-scale and wide ranging negative biological and ecological impacts on a wide range of non-target invertebrates in terrestrial, aquatic, marine and benthic habitats.