Tropospheric ozone is a highly oxidative pollutant with the potential to alter plant metabolism. The direct effects of ozone on plant phenotype may alter interactions with other organisms, such as pollinators, and, consequently, affect plant reproductive success. In a set of greenhouse experiments, we tested whether exposure of plants to a high level of ozone affected their phenological development, their attractiveness to four different pollinators (mason bees, honeybees, hoverflies and bumblebees) and, ultimately, their reproductive success. Exposure of plants to ozone accelerated flowering, particularly on plants that were growing in autumn, when light and temperature cues, that commonly promote flowering, were weaker. Simultaneously, there was a tendency for ozone-exposed plants to disinvest in vegetative growth. Plant exposure to ozone did not substantially affect pollinator preference, but bumblebees had a tendency to visit more flowers on ozone-exposed plants, an effect that was driven by the fact that these plants tended to have more open flowers, meaning a stronger attraction signal. Honeybees spent more time per flower on ozone-exposed plants than on control plants. Acceleration of flower production and the behavioural responses of pollinators to ozone-exposed plants resulted in retained reproductive fitness of plants pollinated by bumblebees, honeybees and mason bees, despite the negative effects of ozone on plant growth. Plants that were pollinated by hoverflies had a reduction in reproductive fitness in response to ozone. In a natural setting, acceleration of flowering by ozone might foster desynchronization between plant and pollinator activities. This can have a strong impact on plants with short flowering periods and on plants that, unlike wild mustard, lack compensatory mechanisms to cope with the absence of pollinator activity in the beginning of flowering.

Recent studies have shown that sub-lethal doses of herbicides may affect plant flowering, however, no study has established a direct relationship between the concentrations of deposited herbicide and plant flowering. Here the aim was to investigate the relationship between herbicide spray drift deposited on non-target plants and plant flowering in a realistic agro-ecosystem setting. The concentrations of the herbicide glyphosate deposited on plants were estimated by measuring the concentration of a dye tracer applied together with the herbicide. The estimated maximal and average deposition of glyphosate within the experimental area corresponded to 30 g glyphosate/ha (2.08% of the label rate of 1440 g a.i./ha) and 2.4 g glyphosate/ha (0.15% label rate), respectively, and the concentrations decreased rapidly with increasing distance from the spraying track. However, there were not a unique relation between distance and deposition, which indicate that heterogeneities of turbulence, wind speed and/or direction can strongly influence the deposition from 1 min to another during spraying. The effects of glyphosate on cumulative flower numbers and flowering time were modelled using Gompertz growth models on four non-target species. Glyphosate had a significantly negative effect on the cumulative number of flowers on Trifolium pratense and Lotus corniculatus, whereas there were no significant effects on Trifolium repens, and a positive, but non-significant, effect on number of flowers on Cichorium intybus. Glyphosate did not affect the flowering time of any of the four species significantly. Lack of floral resources is known to be of major importance for pollinator declines. The implications of the presented results for pesticide risk assessment are discussed.

Neonicotinoids and the closely related insecticide classes sulfoximines and butenolides have recently attracted growing concerns regarding their potential negative effects on non-target organisms, including pollinators such as bees. Indeed, it is becoming increasingly clear that these effects may occur at much lower levels than those considered to be safe for humans. To properly assess the ecological and environmental risks posed by neonicotinoids, appropriate sampling and analytical procedures are needed. Here, we used honey as reliable environmental sampler and developed an unprecedentedly sensitive method based on QuEChERS and UHPLC-MS/MS for the simultaneous determination of the nine neonicotinoids and related molecules currently present on the market (acetamiprid, clothianidin, dinotefuran, flupyradifurone, imidacloprid, nitenpyram, sulfoxaflor, thiacloprid and thiamethoxam). The method was validated and provided excellent levels of precision and accuracy over a wide concentration range of 3–4 orders of magnitude. Lowest limits of quantification (LLOQs) as low as 2–20 pg/g of honey depending on the analytes were reached. The method was then applied to the analysis of 36 honey samples from various regions of the World which had already been analysed for the five most common neonicotinoids (acetamiprid, clothianidin, imidacloprid, thiacloprid and thiamethoxam) in a previous study. This allowed us to determine the long-term stability (i.e. up to 40 months) of these molecules in honey, both at room temperature and −20 °C. We found that the five pesticides were stable over a period of several years at −20 °C, but that acetamiprid and thiacloprid partially degraded at room temperature. Finally, we also measured the levels of dinotefuran, nitenpyram, sulfoxaflor and flupyradifurone and found that 28% of the samples were contaminated by at least one of these pesticides.

The widespread cultivation of transgenic Bt cotton makes assessing the potential effects of this recombinant crop on non-target organisms a priority. However, the effect of Bt cotton on many insects is currently virtually unknown. The plant bug Adelphocoris suturalis is now a major pest of cotton in southern China and the beetle Haptoncus luteolus is one of the most ancient cotton pollinators. We conducted laboratory experiments to evaluate the toxicity of the Bt cotton varieties ZMSJ, which expresses the toxins Cry1Ac and Cry2Ab, and ZMKCKC, which expresses Cry1Ac and EPSPS, on adult A. suturalis and H. luteolus. No significant increase in the mortality of either species was detected after feeding on Bt cotton leaves or pollen for 7 days. Trace amounts of Cry1Ac and Cry2Ab proteins could be detected in both species but in vitro binding experiments found no evidence of Cry1Ac and Cry2Ab binding proteins. These results demonstrate that feeding on the leaves or pollen of these two Bt cotton varieties has no toxic effects on adult A. suturalis or H. luteolus.

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 massive use of plastic has contributed to huge quantities of hazardous refuse at a global scale and represents one of the most prominent issues of the Anthropocene. Microplastics (MPs) have been detected in almost all environments and pose a potential threat to a variety of plant and animal species. Many studies have reported a variety of effects, from negligible to detrimental, of MPs to aquatic organisms. Conversely, much less is known about their effect on terrestrial biota, and particularly on animal behavior and cognition. We assessed the oral toxicity of polyethylene (PE) MPs at three different concentrations (0.5, 5, and 50 mg L⁻¹), and at different timescales (1 day and 7 days of exposure) and tested for their effects on survival, food intake, sucrose responsiveness, habituation to sucrose and appetitive olfactory learning and memory in the honey bee Apis mellifera. We found that workers were not completely unaffected by acute and prolonged ingestion of this polymer. A significant effect of PE on bee mortality was found for the highest concentration but not for lower ones. PE affected feeding behavior in a concentration-dependent manner, with bees consuming more food than controls when exposed to low concentration PE. Regarding our behavioral and cognitive experiments, the high concentration PE was found to affect only bees’ ability to respond consistently to sucrose but not sucrose sensitivity, habituation to sucrose or learning and memory abilities, even for prolonged exposure to PE. While these last results may look somewhat encouraging, we discussed why caution is warranted before ruling out the possibility that PE particles at environmental concentrations are harmful to honey bees.

Tropospheric ozone is a highly oxidative pollutant with the potential to alter plant metabolism. The direct effects of ozone on plant phenotype may alter interactions with other organisms, such as pollinators, and, consequently, affect plant reproductive success. In a set of greenhouse experiments, we tested whether exposure of plants to a high level of ozone affected their phenological development, their attractiveness to four different pollinators (mason bees, honeybees, hoverflies and bumblebees) and, ultimately, their reproductive success. Exposure of plants to ozone accelerated flowering, particularly on plants that were growing in autumn, when light and temperature cues, that commonly promote flowering, were weaker. Simultaneously, there was a tendency for ozone-exposed plants to disinvest in vegetative growth. Plant exposure to ozone did not substantially affect pollinator preference, but bumblebees had a tendency to visit more flowers on ozone-exposed plants, an effect that was driven by the fact that these plants tended to have more open flowers, meaning a stronger attraction signal. Honeybees spent more time per flower on ozone-exposed plants than on control plants. Acceleration of flower production and the behavioural responses of pollinators to ozone-exposed plants resulted in retained reproductive fitness of plants pollinated by bumblebees, honeybees and mason bees, despite the negative effects of ozone on plant growth. Plants that were pollinated by hoverflies had a reduction in reproductive fitness in response to ozone. In a natural setting, acceleration of flowering by ozone might foster desynchronization between plant and pollinator activities. This can have a strong impact on plants with short flowering periods and on plants that, unlike wild mustard, lack compensatory mechanisms to cope with the absence of pollinator activity in the beginning of flowering.

The ecological and economic contributions of pollinator bees to agricultural production have been threatened by the inappropriate and excessive use of pesticides. These pesticides are often applied in areas with ecological peculiarities (e.g., the Neotropical savannah-like region termed as Cerrado) that were not considered during the product development. Here, we conducted field experiments with melon (i.e., Cucumis melo L.) plants cultivated under Brazilian Cerrado conditions and evaluated the impacts of botanical (i.e., neem-based insecticide) and synthetic (i.e., the pyrethroid insecticide deltamethrin and the fungicides thiophanate-methyl and chlorothalonil) pesticides on the flower visitation rates of naturally occurring pollinator bees. Our results revealed that both honey bees (i.e., Apis mellifera L.) and non-Apis bees visited melon flowers and the intensity of bee visitation was moderately correlated with yield parameters (e.g., number of marketable fruits and fruit yield). Pesticide treatments differentially affected bee species. For instance, Plebeia sp. bees were not affected by any pesticide treatment, whereas both A. mellifera and Halictus sp. bees showed reduced visitation intensity after the application of deltamethrin or neem-based insecticides. Fungicide treatment alone did not influence the bee's visitation intensity. Deltamethrin-treated melon fields produced significantly lighter marketable fruits, and the melon yield was significantly lower in melon fields treated with the neem-based insecticide. Thus, our findings with such pollinator bees reinforce the idea that field applications of botanical pesticides may represent as risky as the applications of synthetic compounds, indicating that these alternative products should be submitted to risk assessments comparable to those required for synthetic products.

Plants growing in heavy-metal-rich soils can accumulate metals into their nectar. Nectar chemical composition can alter foraging behavior of floral visitors (including pollinators and floral antagonists) and further affect plant reproductive fitness. The role of nectar heavy metals in deterring pollinators (e.g., shortening foraging time) has been recently studied, but their effects on plant reproduction via changes in behaviors of both pollinators and floral antagonists (e.g., nectar robbers) are less understood. We experimentally manipulated four nectar heavy metals (Zn, Cu, Ni, and Pb) in a native ornamental plant, Hosta ensata F. Maekawa, to investigate the effect of nectar metals on plant reproductive success. We also recorded nectar robbing as well as foraging time and visitation rate of pollinators to assess whether nectar metals could alter the behavior of antagonists and mutualists. Although metals in nectar had no significant direct effects on plant reproduction via hand-pollination, we detected their positive indirect effects on components of female fitness mediated by pollinators and nectar robbers. Matching effects on female plant fitness, nectar robbers responded negatively to the presence of metals in nectar, robbing metal-treated flowers less often. Pollinators spent less time foraging on metal-treated flowers, but their visitation rate to metal-treated flowers was significantly higher than to control flowers. Moreover, pollinators removed less nectar from flowers treated with metals. Our results provide the first direct evidence to date that heavy metals in nectar are capable of deterring nectar robbers and modifying pollinator foraging behavior to enhance plant reproductive fitness.

The worldwide decline of pollinators is of growing concern and has been related to the use of insecticides. Solitary bees are potentially exposed to many insecticides through contaminated pollen and/or nectar. The kinetics of these compounds in solitary bees is, however, unknown, limiting the use of these important pollinators in pesticide regulations. Here, the toxicokinetics (TK) of chlorpyrifos (as Dursban 480 EC), cypermethrin (Sherpa 100 EC), and acetamiprid (Mospilan 20 SP) was studied for the first time in Osmia bicornis females at sublethal concentrations (near LC₂₀ₛ). The TK of the insecticides was analysed in bees continuously exposed to insecticide-contaminated food in the uptake phase followed by feeding with clean food in the decontamination phase. The TK models differed substantially between the insecticides. Acetamiprid followed the classic one-compartment model with gradual accumulation during the uptake phase followed by depuration during the decontamination phase. Cypermethrin accumulated rapidly in the first two days and then its concentration decreased slowly. Chlorpyrifos accumulated similarly rapidly but no substantial depuration was found until the end of the experiment. Our study demonstrates that some insecticides can harm solitary bees when exposed continuously even at trace concentrations in food because of their constant accumulation leading to time-reinforced toxicity.