Sensitivity of bumblebees and solitary bees to neonicotinoids
Direct lethality of neonicotinoids to adult wild bees
Almost all of the studies conducted on the toxicity of neonicotinoids to bees have been conducted on honeybees, Apis mellifera. Fourteen studies conducted up to 2010 were reviewed in a meta-analysis by Cresswell (2011) who concluded that for acute oral toxicity imidacloprid has a 48-h LD50 = 4.5 ng/bee. The EFSA studies (2013a, b, c) reviewed existing studies for acute oral toxicity up to 2013, including both peer-reviewed studies and also private studies that are not in the public domain (summarised in Godfray et al. 2014). These analyses produced LD50s of 3.7 ng/bee for imidacloprid, 3.8 ng/bee for clothianidin and 5.0 ng/bee for thiamethoxam. Equivalent LD50s for acute contact have also been calculated by EFSA (2013a, b, c) for honeybees to be 81 ng/bee for imidacloprid, 44 ng/bee for clothianidin and 24 ng/bee for thiamethoxam.
However, the EFSA reports highlighted a knowledge gap for the effects of neonicotinoids on bees other than honeybees. Arena and Sgolastra (2014) conducted a meta-analysis comparing the sensitivity of bees to pesticides relative to the sensitivity of honeybees. This analysis combined data from 47 studies covering 53 pesticides from six chemical families with a total of 150 case studies covering 18 bee species (plus A. mellifera). Arena and Sgolastra calculated a sensitivity ratio R between the lethal dose for species a (A. mellifera) and for species s (other than A. mellifera), R = LD50
a/LD50
s. A ratio of over 1 indicates that the other bee species is more sensitive to the selected pesticides than A. mellifera and vice versa. There was high variability in relative sensitivity ranging from 0.001 to 2085.7, but across all pesticides a median sensitivity of 0.57 was calculated, suggesting that A. mellifera was generally more sensitive to pesticides than other bee species. In the vast majority of cases (95%), the sensitivity ratio was below 10.
Combining data for all neonicotinoids (acetamiprid, imidacloprid, thiacloprid and thiamethoxam) and for both acute contact and acute oral toxicity, nine studies covering nine bee species (plus A. mellifera) were found. These studies showed a median sensitivity ratio of 1.045 which is the highest median value of all the analysed pesticide chemical families. The most relatively toxic neonicotinoids to other bees were the cyano-substituted neonicotinoids acetamiprid and thiacloprid as these exhibit lower toxicity to honeybees than the nitro-substituted neonicotinoids imidacloprid and thiamethoxam.
Selecting pesticides covered by the moratorium (excluding acetamiprid and thiacloprid and including fipronil) and including both acute contact and acute oral toxicity, 12 studies covering 10 bee species (plus A. mellifera) were found. These studies showed a median sensitivity ratio of 0.957 which is close to the calculated sensitivity ratio for all neonicotinoids. The greatest discrepancy between honeybees and other bees was found for stingless bees (Apidae: Meliponini). The effect of acute contact of fipronil on Scaptotrigona postica (24-fold greater), of acute contact of fipronil on Melipona scutellaris (14-fold greater) and of acute contact of Thiacloprid on Nannotrigona perilampoides (2086-fold) were the only three cases with a sensitivity ratio of over 10. Stingless bees are predominantly equatorial with the greatest diversity found in the neotropics. No species are found in Europe (Nieto et al. 2014). In contrast, studies on B. terrestris consistently report a lower sensitivity ratio between 0.005 and 0.914, median 0.264. B. terrestris is widespread in Europe and is the most commonly used non-Apis model system for assessing the effects of neonicotinoids on wild bees (see the “Sublethal effects of neonicotinoids on wild bees” section). Differences in bee body weight have been proposed to explain these differences, with sensitivity to pesticides inversely correlated with body size (Devillers et al. 2003). However, this has not been consistently demonstrated and other mechanisms have been suggested such as species level adaptation to feeding on alkaloid-rich nectar (Cresswell et al. 2012) and differential abilities to clear neonicotinoid residues from their bodies (Cresswell et al. 2014). With the limited data available, Arena and Sgolastra could not comment on the strength of these claims.
Spurgeon et al. (2016) calculated various toxicity measures of clothianidin on honeybees, the bumblebee species B. terrestris and the solitary bee species O. bicornis. Acute oral toxicity 48-, 96- and 240-h LD50s for honeybees were 14.6, 15.4 and 11.7 ng/bee respectively. For B. terrestris, the corresponding values were 26.6, 35 and 57.4 ng/bee respectively. For O. bicornis, the corresponding values were 8.4, 12.4 and 28.0 ng/bee respectively. These findings are generally in line with the findings of Arena and Sgolastra, with B. terrestris less sensitive than A. mellifera at all time points and O. bicornis less sensitive at 240 h.
Sgolastra et al. (2016) calculated relative sensitivity to clothianidin to these same three species over a range of time periods from 24 to 96 h. The highest LD50 values were obtained after 24 h for A. mellifera and B. terrestris and after 72 h for O. bicornis. At these time points, O. bicornis was the most sensitive of the three species, with LD50 measurements of 1.17 ng/bee and 9.47 ng/g, compared to 1.68 ng/bee and 19.08 ng/g for A. mellifera and 3.12 ng/bee and 11.90 ng/g for B. terrestris. These results are in line with the values calculated by Spurgeon et al. (except for the 240-h values), with decreasing sensitivity in the order of O. bicornis > A. mellifera > B. terrestris. Together, these studies support the position that small-bodied species show greater sensitivity to neonicotinoids.
Around 2000 bee species are known from Europe. The biology, behaviour and ecology of each of these species differ from those of honeybees. Consequently, extrapolating from the limited toxicological data available for 19 bee species to the effects of neonicotinoids on the wider European fauna is fraught with difficulties given the wide variation in relative sensitivity. Current data suggests that wild bees are equally to slightly less sensitive to neonicotinoids compared to honeybees when considering direct mortality. However, care must be taken when considering individual bee species, genera and families, as different taxonomic groups may show consistently different individual-level sensitivity. Most European wild bees are smaller than honeybees, and there is the potential for them to be more sensitive on a nanogram per bee basis. In general, continuing to use honeybee neonicotinoid sensitivity metrics is likely to be a reasonable proxy measure for the direct sensitivity of the wild bee community to neonicotinoids (Arena and Sgolastra 2014).
Sublethal effects of neonicotinoids on wild bees
In 2013, a number of studies looking at sublethal effects of neonicotinoids were available, predominantly using honeybees as a model organism in laboratory conditions. Blacquière et al. (2012) reviewed studies on neonicotinoid side effects on bees published between 1995 and 2011 with a specific focus on sublethal effects. The authors found that whilst many laboratory studies described lethal and sublethal effects of neonicotinoids on the foraging behaviour and learning and memory abilities of bees, no effects were observed in field studies at field-realistic dosages. Two major studies that substantially contributed towards the initiation and subsequent implementation of the EU neonicotinoid moratorium were published after this review in 2012.
Henry et al. (2012) gave honeybee workers an acute dose of 1.34 ng of thiamethoxam in a 20 μL sucrose solution, equivalent to 27% of the LD50 (see the “Direct lethality of neonicotinoids to adult wild bees” section), then released them 1 km away from their nests and measured their return rate. Dosed bees were significantly less likely to return to the nest than control bees. Whitehorn et al. (2012) exposed B. terrestris colonies to two levels of neonicotinoid-treated pollen (6 and 12 ng/g plus control) and nectar (0.7 and 1.4 ng/g plus control) in the laboratory for 2 weeks before moving them outdoors to forage independently for 6 weeks, aiming to mimic a pulse exposure that would be expected for bees foraging on neonicotinoid-treated oilseed rape. Bees in the two neonicotinoid treatments grew significantly more slowly and had an 85% reduction in the number of new queens produced when compared to control colonies.
Both of these studies have been criticised for using neonicotinoid concentrations greater than those wild bees are likely to be exposed to in the field (see Godfray et al. 2014; Carreck and Ratnieks 2014). The 1.34 ng of thiamethoxam in a 20 μL sucrose solution used by Henry et al. is a concentration of 67 ng/g. Taking maximum estimated concentrations of thiamethoxam in oilseed rape nectar of 2.72 ng/g (see the “Risk of exposure from pollen and nectar of treated flowering crops” section), a honeybee would have to consume 0.49 g of nectar to receive this dose. Honeybees typically carry 25–40 mg of nectar per foraging trip, equivalent to 0.025–0.040 g, some 10% of the volume necessary to receive a dose as high as the one used by Henry et al. Moreover, as honeybee workers regurgitate this nectar at the hive, the total dose consumed is likely to be a fraction of the total amount carried. Consequently, it is extremely unlikely that the findings of Henry et al. are representative of a real-world situation.
The pollen and nectar concentrations used by Whitehorn et al. are much closer to field-realistic levels with the lower treatment within maximum estimated concentrations of imidacloprid in oilseed rape pollen and nectar (see the “Risk of exposure from pollen and nectar of treated flowering crops” section). However, the experimental setup, where bumblebees had no choice but to consume treated pollen and nectar, has been criticised as unrealistic, as in the real-world alternative, uncontaminated forage sources would be available. Studies that have measured residues in both crop and wildflower pollen and have assessed the origin of bee-collected pollen (see “Risk of exposure from and uptake of neonicotinoids in non-crop plants” section) have recorded neonicotinoid concentrations of between 0.84 and 27.0 ng/g in wild bee-collected pollen where a substantial proportion of this pollen is collected from crop plants during their period of peak flowering. Pollen extracted from bumblebee nests contained neonicotinoid concentrations of 6.5 ng/g in urban areas and 21.2 ng/g in rural areas during the peak flowering period of oilseed rape, though the number of nests sampled (three and five) were low. However, other studies measuring levels in pollen taken directly from bumblebees found concentrations of <1 ng/g, so there is still a lack of clarity surrounding true levels of neonicotinoid exposure for wild bumblebees. On the basis of these described concentrations, the results of Whitehorn et al. are likely to be closer to real-world conditions than the findings of Henry et al.
Post April 2013, much work on sublethal effects of neonicotinoids on bees has been carried out on individual honeybees and honeybee colony fitness metrics, such as colony growth, overwintering success and the production of sexuals. This work is beyond the scope of this review, but important recent publications include Pilling et al. (2013), Cutler et al. (2014), Rundlöf et al. (2015) and Divley et al. (2015) who all found limited to negligible impacts of neonicotinoids at the colony level. See also Cresswell (2011) for a meta-analysis of 13 laboratory and semi-field studies conducted before 2011. Various authors note that interpreting the findings of studies on honeybees to wild bees is fraught with difficulty, given the differing size of individual bees and the social behaviour of honeybees that gives rise to colonies containing many thousands of workers.
Impact on colony growth and reproductive success
Several authors have investigated the effects of neonicotinoids on bumblebees using micro-colonies. These are small groups of worker bumblebees that are taken from a queenright colony and isolated in a new nest box. These workers, lacking a queen, will begin to rear their own male offspring. As such, micro-colonies are useful for generating a large sample size for investigating pesticide impacts on bee mortality and larval rearing behaviour and reproductive success.
Elston et al. (2013) fed micro-colonies of three B. terrestris workers a ‘field-realistic’ dose of 1 ng/g thiamethoxam and a ‘field-maximum’ dose of 10 ng/g in both pollen paste and sugar solution for a 28-day period. Micro-colonies from both thiamethoxam treatments consumed significantly less sugar solution than control colonies. There was no impact on worker mortality, but colonies fed 10 ng/g thiamethoxam had reduced nest-building activity and produced significantly fewer eggs and larvae, with the 10 ng/g thiamethoxam treatment the only one to produce no larvae over the 28-day experimental period.
Laycock et al. (2014) fed micro-colonies of four B. terrestris workers thiamethoxam-treated sugar solution at a range of concentrations up to 98 ng/g. Pollen was not treated with thiamethoxam. Sugar solution consumption was significantly reduced at the 39 and 98 ng/g treatments. Worker mortality was only increased at the highest dose of 98 ng/g. Worker oviposition failure was only significantly higher at the 39 and 98 ng/g treatments, with no significant differences seen between the lower-concentration treatments between 0 and 16 ng/g.
The findings of these two studies are generally in line with pre-2013 knowledge. Mommaerts et al. (2010) exposed B. terrestris micro-colonies to sugar solution treated with thiamethoxam concentrations of up to 100 ng/g. Whilst the 100 ng/g level reduced brood production, the 10 ng/g treatment had no detectable effect. The difference between the findings of Elston et al. and Laycock et al. may partially be explained by the fact that Elston et al. treated pollen with thiamethoxam as well as sugar solution. Laycock et al. confirm that concentrations of 98 ng/g increase worker mortality, but as such concentrations are not usually encountered in the field, this is of limited relevance.
Scholer and Krischik (2014) exposed greenhouse queenright colonies of B. impatiens to imidacloprid- and clothianidin-treated sugar syrup at concentrations of 0, 10, 20, 50 and 100 ng/g for 11 weeks. Queen mortality was significantly increased at 6 weeks for the 50 and 100 ng/g treatments, and at 11 weeks for the 20 ng/g treatment for both clothianidin and imidacloprid. Surprisingly, no significant impact was found on numbers of workers or new queens produced, though this was in part because very low numbers of new queens were produced across all treatments (average of four per colony). Colonies in treatments above 10 ng/g imidacloprid and 20 ng/g of clothianidin gained significantly less weight over the course of the study. Neonicotinoid concentrations of 20 ng/g and above are very high and are unlikely to be consistently encountered by bees for prolonged periods of times under real-world conditions. As a result, queen mortality in the real world is unlikely to be significantly affected by currently observed neonicotinoid concentrations.
Several field studies have also been published since 2013 that investigate the impact of neonicotinoid-treated mass flowering crops on wild bee colony growth and reproductive success. Cutler and Scott-Dupreee (2014) placed B. impatiens colonies adjacent to maize fields during pollen shed in Ontario, Canada. Four neonicotinoid-treated conventional and four untreated organic fields were used. Colonies were placed out adjacent to each field on the first day of major pollen shed. Colonies were left for 5–6 days and then transported to an area of semi-natural habitat for 30–35 days, after which they were frozen. Colonies placed next to treated maize produced significantly fewer workers than those placed next to organic farms. All other metrics (colony weight, honey and pollen pots, brood cells, worker weight, male and queen numbers and weights) were not significantly different. Bumblebees collected less than 1% of their pollen from maize (“Risk of exposure from and uptake of neonicotinoids in non-crop plants” section) and neonicotinoid residues in collected pollen were low, at 0.4 ng/g from bees foraging adjacent to treated fields and below the LOD for bees adjacent to organic fields. Given that it is well known that bumblebees collect very low volumes of maize pollen, the relevance of this study is unclear.
Rundlöf et al. (2015) conducted an extensive field trial of the effects of clothianidin-treated oilseed rape on wild bees. Sixteen oilseed rape fields separated by at least 4 km were selected across southern Sweden and were paired on the basis of similar landscape composition. In each pair, one of the fields was randomly selected to be sown with oilseed rape treated with 10 g clothianidin/kg of seed and the other field was sown without a neonicotinoid seed treatment. Twenty-seven cocoons of the solitary bee O. bicornis (15 male, 12 female) were placed out alongside each field a week before the oilseed rape began to flower, and six colonies of B. terrestris were placed alongside each field on the day the oilseed rape began to flower. The O. bicornis placed adjacent to treated oilseed rape showed no nesting behaviour and did not initiate brood cell construction. O. bicornis adjacent to untreated fields showed nesting behaviour in six of the eight fields studied. The reasons for these differences in nest initiation are unclear, and it is difficult to draw firm conclusions with a small sample size. Bumblebees placed next to treated oilseed rape showed reduced colony growth and reproductive output. Bumblebee colonies were collected and frozen when new queens began to emerge, with this happening between the 7th of July and 5th of August depending on each colony. The number of queen and worker/male cocoons present was counted. At the point of freezing, colonies placed next to treated oilseed rape fields had significantly fewer queen and worker/male cocoons present.
Sterk et al. (2016) performed a similar field experiment to Rundlöf et al. Two 65 km2 areas in northern Germany were selected in which the only flowering crops comprised winter-sown oilseed rape. In one area, the oilseed rape was treated with the same seed coating used by Rundlöf et al. of 10 g clothianidin/kg seed. The other area was an untreated control. In each area, 10 B. terrestris colonies were placed at each of six localities. Colonies were left adjacent to oilseed rape between April and June, covering its main flowering period. After this, the colonies were moved to a nature reserve. No differences were found in colony weight growth, number of workers produced or reproductive output as measured by the production of new queens.
That these two field studies using the same neonicotinoid seed dressing found markedly different results is interesting. The major difference is that whilst Rundlöf et al. used spring-sown oilseed rape, Sterk et al. used winter-sown oilseed rape. The length of time between sowing and peak flowering is much greater for winter-sown oilseed rape (mid-August to May) than for spring-sown oilseed rape (April/May to mid-June). As such, there is more time for neonicotinoids to leach into soil and water for winter-sown oilseed rape, reducing the amount of active ingredient available to be taken up by the crop. This may explain some of the order of magnitude differences in neonicotinoid concentrations in pollen collected from the two crops (“Risk of exposure from and uptake of neonicotinoids in non-crop plants” section) and the difference in reported colony growth and number of reproductives produced. An additional difference is that in the Sterk et al. study, colonies were moved to a nature reserve consisting of forests, lakes and heaths after the flowering period of oilseed rape ended. The quality of available forage at this nature reserve is likely to have been of both a higher quality and a higher quantity than what was available in a conventional agricultural landscape and is not typical of the experience of a bumblebee colony located in such a landscape that will have to continue foraging there after crops such as oilseed rape cease flowering. In addition, Sterk et al. had only one treated and one control area, so there is no true site level replication, as opposed to Rundlöf et al. who used eight treated and eight control fields. These differences in experimental design should be taken into account when considering why the studies produced such different results.
One of the studies conducted in response to the results of Henry et al. (2012) and Whitehorn et al. (2012) was produced by FERA (2013). It consisted of a field trial with bumblebee colonies placed out adjacent to oilseed rape treated with either clothianidin, imidacloprid or an untreated control. Colonies were allowed to forage freely for 6–7 weeks whilst the oilseed rape flowered and then were moved to a non-agricultural area to continue developing. The study was ultimately not published in a peer-reviewed journal, but it came to the conclusion that there was no clear relationship between bumblebee colony success and neonicotinoid concentrations. Goulson (2015) reanalysed the FERA data using linear models and retaining two colonies excluded in the original study as outliers, but which do not meet the statistical definition of this term. This reanalysis found that the concentration of clothianidin in nectar and the concentration of thiamethoxam in pollen significantly negatively predicted both colony weight gain and production of new queens.
Only one study is available that looked at the impact of neonicotinoids on the reproductive success of a solitary bee in controlled conditions. Sandrock et al. (2014) established laboratory populations of O. bicornis, a solitary stem nesting bee. Bees were fed on sugar solution treated with 2.87 ng/g thiamethoxam and 0.45 ng/g clothianidin along with untreated pollen. There was no impact of neonicotinoids on adult female longevity or body weight. However, treated bees completed 22% fewer nests over the course of the experiment. Nests completed by treated bees contained 43.7% fewer total cells, and relative offspring mortality was significantly higher, with mortality rates of 15 and 8.5% in the treated and untreated groups respectively. Overall, chronic neonicotinoid exposure resulted in a significant reduction in offspring emergence per nest, with treated bees producing 47.7% fewer offspring. These results suggest that exposure to these low-level, field-realistic doses of neonicotinoids (<3.5 ng/g) did not increase adult mortality but did have sublethal impacts on their ability to successfully build nests and provision offspring. However, it is important to note that this study had no true replication, and thus, the results should be interpreted with considerable caution.
Overall, the studies produced since 2013 are generally in line with existing knowledge at this point but have advanced our knowledge in several key areas. Laboratory studies have continued to demonstrate negative effects of neonicotinoids on bumblebee reproductive output at generally high concentrations, with the lowest sublethal effects on reproductive output detected at 10 ng/g. Field studies using bumblebees demonstrate that exposure to neonicotinoid-treated flowering crops can have significant impacts on colony growth and reproductive output depending on the levels exposed to, with crop flowering date relative to sowing and availability of uncontaminated forage plants likely to explain variation in the detected residues between the available studies. Our understanding of the impact on solitary bees is much improved with the findings of Sandrock et al. (2014) suggesting substantial impacts on solitary bee reproductive output at field-realistic concentrations of 3.5 ng/g. Field studies demonstrating this under real-world conditions are limited with the work of Rundlöf et al. (2015) suffering from no nest-building activity at the neonicotinoid treatment sites.
Impact on foraging efficiency
In 2013, a limited amount was known about how neonicotinoids affected the foraging behaviour of individual bees, and whether this affected colony level fitness. Gill et al. (2012) exposed B. terrestris colonies to 10 ng/g imidacloprid in sugar solution in the nest for a period of 4 weeks. Colonies were housed indoors, but access tubes allowed them to forage freely outdoors. Imidacloprid-exposed colonies grew more slowly, but there were substantial effects on worker foraging behaviour. Compared to controls, imidacloprid-treated colonies had more workers initiating foraging trips, workers brought back smaller volumes of pollen on each successful trip and successful pollen foraging trips were of a significantly longer duration. Treated workers also collected pollen less frequently, with 59% of foraging bouts collecting pollen versus 82% for control workers, a decline of 28%. The authors conclude that exposure to imidacloprid at these concentrations significantly reduced the ability of bumblebee workers to collect pollen in the field. The reduced ability to collect pollen resulted in imidacloprid-treated colonies collecting less pollen than control colonies, subsequently resulting in reduced growth through pollen limitation. Since the publication of this paper, several new studies assessing neonicotinoid impacts on the foraging behaviour of bumblebees have been published.
Feltham et al. (2014) exposed B. terrestris colonies to sugar solution treated with 0.7 ng/g and pollen treated with 6 ng/g of imidacloprid for 2 weeks. These sugar solution concentrations were an order of magnitude lower than the 10 ng/g used by Gill et al. (2012). Colonies were then placed out in an urban area in Scotland. The foraging workers from each nest were then monitored for a further 4 weeks. There was no difference in the length of time spent collecting nectar or the volume of nectar collected between workers from treated and control colonies. However, treated workers collected significantly less pollen, bringing back 31% less pollen per time unit to their colonies. Treated workers also collected pollen less frequently, with 41% of foraging bouts collecting pollen versus 65% for control workers, a decline of 23%.
Gill and Raine (2014) performed a similar experiment to Gill et al. (2012) where B. terrestris colonies were exposed to sugar solution treated with 10 ng/g of imidacloprid whilst also having access to forage freely outside. Colonies and individual worker bumblebees were studied over a 4-week period. In common with their previous findings (Gill et al. 2012), imidacloprid-treated workers initiated significantly more foraging trips across all 4 weeks of the experiment. The authors note that this is likely driven by an acute individual-level response in the first weeks (neonicotinoids acting as a neural partial agonist, increasing desire to forage) and by a chronic colony-level response in the latter part of the experiment, with treated colonies allocating a higher proportion of workers to pollen collection. Pollen foraging efficiency of treated workers decreased as the experiment progressed with the smallest collected pollen loads recorded in week 4, suggesting a chronic effect of imidacloprid on pollen foraging ability. It is not clear whether this is as a result of individual performance deteriorating, or new emerging workers having been exposed for a greater period of time.
Stanley et al. (2015) exposed B. terrestris colonies to 2.4 or 10 ng/g thiamethoxam-treated sugar solution for 13 days. Colonies were then moved to pollinator exclusion cages where they were allowed to forage freely on two varieties of apple blossom. Bees from colonies exposed to 10 ng/g spent longer foraging, visited fewer flowers and brought back pollen on a lower proportion of foraging trips compared to bees from control colonies. Stanley and Raine (2016) also exposed B. terrestris colonies to 10 ng/g thiamethoxam sugar solution for a 9- to 10-day period. At this point, colonies were moved to a flight arena provisioned with two common bird’s-foot trefoil Lotus corniculatus plants and one white clover Trifolium repens plant. Worker bees were individually released, and their interaction with the flowers was recorded. Significantly more treated workers displayed pollen-foraging behaviour compared to control workers. However, control workers learnt to handle flowers efficiently after fewer learning visits.
Arce et al. (2016) placed B. terrestris nests out in an area of parkland for a 5-week period whilst also supplying them with sugar solution treated with 5 ng/g of clothianidin. The volume of sugar solution provided was estimated to be half that which colonies typically consume over the course of the experiment. No pollen was provided, so workers had to forage for this and to make up the shortfall in nectar resources. In contrast to the previous papers, only subtle changes to patterns of foraging activity and pollen collection were detected. There was no clear difference in colony weight gain between treatments or number of brood individuals. However, by the end of the experiment, treated colonies contained fewer workers, drones and gynes when compared with control colonies.
Switzer and Combes (2016) studied the impact of acute imidacloprid ingestion on sonicating behaviour of B. impatiens. Sonicating is a behaviour whereby a bumblebee lands on a flower and vibrates loudly to shake pollen loose from anthers. Bumblebee workers were fed a dose of 0, 0.0515, 0.515 or 5.15 ng of imidacloprid in 10 μL of sugar solution. These are equivalent to concentrations of 0, 5.15, 51.5 and 515 ng/g, with the highest volume consumed equivalent to 139% of the honeybee LD50, a moderate proxy for bumblebees (see “Direct lethality of neonicotinoids to adult wild bees” section). Bees were then allowed to forage from tomato Solanum lysopersicum plants, and sonicating behaviour was observed. At the lowest dose of 0.0515 ng of imidacloprid, no impact was found on wingbeat frequency, sonication frequency or sonication length. No analysis could be made for higher doses, as bees in these treatments rarely resumed foraging behaviour after ingesting imidacloprid. Given the neonicotinoid concentrations used in this study and the lack of observed sonicating behaviour at higher doses, it is difficult to draw many conclusions other than that high levels of exposure may impair bumblebee pollen foraging behaviour.
Overall, these studies suggest that exposure to neonicotinoids in nectar at concentrations of between 0.7 and 10 ng/g can have sublethal effects on the ability of bumblebees to collect pollen at both the individual and colony levels. This shortfall in pollen and subsequent resource stress is a plausible mechanism to explain diminished colony growth and production of sexuals in the absence of increased direct worker mortality. Given that concentrations as high as 10 ng/g are at, but within, the upper limit of what bumblebees are likely to experience in the field (“Risk of exposure from pollen and nectar of treated flowering crops” and “Risk of exposure from and uptake of neonicotinoids in non-crop plants” sections), it is likely that wild bumblebees exposed to neonicotinoids in contemporary agricultural environments suffer from a reduced ability to collect pollen, with a subsequent impact on their reproductive output.
Impact on bee immune systems
Bee diseases (including both parasites and pathogens) have been implicated as the major factor affecting managed honeybee colony survival in recent years (vanEngelsdorp et al. 2010). Whilst most evidence for the negative effects of diseases comes from studies of honeybees, most diseases can affect a wide range of bee species. For example, the microsporidian parasite Nosema ceranae, originates in Asia but has spread around the world during the last 20 years, probably as a result of the international trade in honeybees (Klee et al. 2007). N. ceranae has now been detected in four different genera of wild bees (Bombus, Osmia, Andrena, Heriades) across Europe and the Americas (see Goulson et al. 2015). The spread of diseases between wild and managed bees can occur at shared flowering plants (Graystock et al. 2015).
Sánchez-Bayo et al. (2016) reviewed evidence that linked the use of neonicotinoids to the incidence and severity of bee diseases. Prior to 2013, several studies demonstrated a link between neonicotinoid exposure and increased susceptibility to diseases in honeybees (Vidau et al. 2011; Pettis et al. 2012). Exposure of honeybees infected with N. ceranae to imidacloprid reduced their ability to sterilise the brood, increasing the spread of N. ceranae within the colonies (Alaux et al. 2010). In addition, exposure to sublethal doses of imidacloprid or fipronil increased honeybee worker mortality due to a suppression of immunity-related genes (Aufauvre et al. 2012). Di Prisco et al. (2013) found that sublethal doses of clothianidin adversely affected honeybee antiviral defences. By enhancing the transcription of the gene encoding a protein that inhibits immune signalling activation, the neonicotinoid pesticides reduce immune defences and promote the replication of deformed wing virus in honeybees bearing covert viral infections. At the field level, a positive correlation is found between neonicotinoid treatment and Varroa mite infestation and viral load of honeybee colonies (Divley et al. 2015; Alburaki et al. 2015). No studies are available that measure the impact of neonicotinoids on the immune systems of wild bees or on the incidence of diseases in wild bees in conjunction with neonicotinoid usage. However, given that wild bees share a very similar nervous and immune system, it is highly likely that neonicotinoids will have similar effects, increasing wild bee susceptibility to parasites and pathogens.
Population-level effects of neonicotinoids on wild bees
Nothing was known about the population level effects of neonicotinoids on wild bees in 2013. As a managed domesticated species, population trends are available for honeybees, but no such data are available for wild bees. One study has attempted to investigate the impact of neonicotinoids on wild bee population trends. Woodcock et al. (2016) used an incidence dataset of wild bee presence in 10 × 10 km grid squares across the UK. The dataset is composed of bee sightings by amateur and professional entomologists and is probably the most complete national bee distribution database currently in existence. Sixty-two wild bee species were selected, and their geographic distance and persistence over an 18-year period between 1994 and 2011 was calculated. Neonicotinoid seed-treated oilseed rape was first used in the UK in 2002, and so the authors calculated spatially and temporally explicit information describing the cover of oilseed rape and the area of this crop treated with neonicotinoids. The 62 species were split into two groups—species that foraged on oilseed rape (n = 34) and species that did not (n = 28). Species persistence across this time period was then compared with expected neonicotinoid exposure. Over the 18-year period, wild bee species persistence was significantly negatively correlated with neonicotinoid exposure for both the foraging and non-foraging groups, with the effect size three times larger for the oilseed rape foraging group.
The characterisation of bees as foragers or non-foragers has one major problem. Many species of bees are obligately parasitic on other bees and do not forage for their own pollen. Some parasitic bees were included in the oilseed rape forager category (n = 2), and some in the non-forager category (n = 12) based on observed nectar visits from a previous study. Some of the parasitic bees in the non-forager group are parasitic on bees included in the forager group (n = 10/28). Given that these species are highly dependent on their host’s abundance, this classification does not make ecological sense. A decline due to a decline in their host or because of increased direct mortality cannot be separated, introducing an additional confounding issue into the analysis. In addition, given the presence of neonicotinoids in wild plants adjacent to agricultural areas (“Risk of exposure from and uptake of neonicotinoids in non-crop plants” section), the amount applied to oilseed rape is not necessarily a true measure of actual neonicotinoid exposure for wild bees.
Overall, the study suggests that bee species were more likely to disappear from areas with a high exposure to neonicotinoids as measured by the amounts applied as seed dressings to oilseed rape, and that this trend was more pronounced for species known to forage on oilseed rape.
Sensitivity of butterflies and moths to neonicotinoids
Pisa et al. (2015) reviewed the existing literature on the impact of neonicotinoids on butterflies and moths (Lepidoptera). In contrast to bees, very few comparative toxicity tests have been conducted for butterflies. Most existing studies have compared butterfly abundance and diversity on organic versus conventional farms. Organic farms host a greater diversity of species, but the specific reasons for this cannot be isolated. For example, the relative importance of herbicide use that reduces the abundance of larval food and adult nectar plants versus direct mortality or sublethal stress from pesticides is unknown.
Most available toxicological studies looking at the sensitivity of Lepidoptera to neonicotinoids and fipronil have been conducted on 32 species of moths from nine families that are pests of crops (Pisa et al. 2015). There is considerable variation in reported sensitivities between species, with the susceptibility to acetamiprid of two cotton pests differing almost 3-fold (LC50 = 11,049 and 3798 ppm). There is also variation between different stages of larval development, with first instar caterpillars more than 100 times as sensitive as fifth instar caterpillars with a LC50/LC90 of 0.84/1.83 and 114.78/462.11 ppm respectively. Botías et al. (2016) listed LC50 values for three moth species that are agricultural crop pests, with 24-h LC50 values between 2400 and 186,000 ppb clothianidin. These levels are generally very high, and there are multiple examples of neonicotinoid resistance in wild populations (see Pisa et al. 2015). Because many of the studied moths species are pests of major crops, they have been exposed to multiple pesticides over many generations in recent decades, and their sensitivity to neonicotinoids may not necessarily be representative of non-pest wild Lepidoptera species.
Since 2013, few studies looking at the sensitivity of wild Lepidoptera to neonicotinoids are available. Pecenka and Lundgren (2015) assessed the lethality of clothianidin to caterpillars of monarch butterflies Danaus plexippus. First instar caterpillars were fed treated leaves for a 36-h period. A LC50 of 15.63 ng/g was calculated. In addition, sublethal effects on growth were measured at 0.5 ng/g with first instar larvae taking longer to develop, having reduced body length and lower weight. These differences did not extend into the second instar. Yu et al. (2015) fed second instar silkworm Bombyx mori caterpillars leaves treated with imidacloprid and thiamethoxam for a 96-h period. They calculated LC50 values of 1270 ng/g for imidacloprid and 2380 ng/g for thiamethoxam. This wide range of reported tolerances for a limited number of ecologically different species means that thorough assessment of butterfly and moth sensitivity to neonicotinoids is difficult. Much more research is required in this area.
Whilst there is a paucity of toxicological data on wild butterflies and moths, two recent studies have used long-term butterfly population datasets to assess the relative impact of neonicotinoid usage in agricultural areas. Gilburn et al. (2015) used data from the UK butterfly monitoring scheme. The data consists of butterfly counts from a wide variety of habitats, and the period studied was 1984–2012, a more extensive time period than that used for UK wild bees by Woodcock et al. (2016, Section 3.1.3) in order to include a 10-year period before the introduction of neonicotinoids onto British farmland. Seventeen UK butterfly species were selected that are predominantly generalists and are found in a wide range of habitats including agricultural habitats. The area of the UK treated with neonicotinoids and a range of temperature and weather variables were included in the model, as local climatic conditions are a very important factor impacting butterfly populations. In line with expectations, summer temperature was significantly positively and spring rainfall significantly negatively correlated with the butterfly population indexes. Neonicotinoid usage was also significantly negatively associated with butterfly population indices after controlling for the effects of weather. The pattern of association varied between butterfly species, but most (14 out of 17) had a negative association. In the most recent time period between 2000 and 2009 when neonicotinoid usage was at its highest, 15 of the 17 studied species showed a negative population trend.
Forister et al. (2016) conducted a similar analysis on Californian lowland butterfly populations. Butterflies have been monitored continuously with biweekly walks at four sites in a region of northern California since 1972, 1975 and 1988 depending on the individual site. These sites are situated across a land gradient that includes arable, semi-natural and urban habitats. The data were used to examine the impact of annual neonicotinoid input and other factors such as summer temperature and land use change.
A substantial decline in butterfly species richness was seen from 1997 onwards, with 1997 being the breakpoint identified by the statistical models. Neonicotinoid usage in the region began in 1995 and has increased since that point. Neonicotinoid use was significantly negatively correlated with butterfly species richness, with smaller-bodied butterflies showing the strongest negative correlation.
Both of these analyses are strictly correlational, and neonicotinoid usage may simply be a proxy measurement for some other factor that is driving declines. Gilburn et al. note that if habitat deterioration and loss of food plants is the main cause of butterfly declines, and agricultural intensification is playing a key role in this habitat deterioration, then levels of neonicotinoid usage might be acting as a proxy for agricultural intensification and therefore habitat deterioration. Thus, neonicotinoid usage could be responsible for driving butterfly declines or alternatively it could provide the first useful quantifiable measure of agricultural intensification that strongly correlates with butterfly population trends. As most of the UK butterfly monitoring scheme survey areas are not directly on agricultural land, Gilburn et al. suspect that it is the transport of neonicotinoids into the wider environment (“Risk of exposure from and uptake of neonicotinoids in non-crop plants” section) and farmed areas acting as population sinks that is driving the declines of butterflies, rather than neonicotinoid use acting as a proxy for agricultural intensification. No data are available to assess this hypothesis.
Overall, recent studies have demonstrated that Lepidoptera show a wide range of tolerances to ingested neonicotinoids in their larval stages. No data are available on sensitivity to neonicotinoids ingested during the adult stage, for example from crop plant nectar. Two correlational studies using long-term datasets show a strong association between neonicotinoid use and declines in butterfly abundance and species richness, though more laboratory and field studies are required to establish the exact mechanism causing this decline.
Sensitivity of other terrestrial invertebrates to neonicotinoids
Most available studies that have assessed neonicotinoid sensitivity for insect species have focused on pest species of economically important crops. Pisa et al. (2015) reviewed existing literature on the impacts of neonicotinoids on other terrestrial invertebrates, and Botías et al. (2016) presented a summary on reported LC50s for 24 species of insects across four orders (Hymenoptera, Lepidoptera, Hemiptera and Coleoptera) from studies conducted between 1996 and 2015. Pisa et al. (2015) review found no post-2013 research on the effects of neonicotinoids on Neuroptera, Hemiptera and Syrphidae (hoverflies).
Four studies are available that have looked at the impact of neonicotinoids on ants. Galvanho et al. (2013) treated Acromyrmex subterraneus leafcutter ants with imidacloprid to investigate impacts on grooming, an important behaviour for limiting the spread of fungal pathogens. Workers were treated with 10, 20 or 40 ng/insect imidacloprid. Only workers with a head capsule of 1.6–2.0 mm in width were selected. This is a large size relative to most species of ants in the world. At this size, individual ants would weigh around 10–20 mg, giving a concentration of 10–40 ng active ingredient per 0.015 g of ant, or 666.7–2666.7 ng/g. The lowest dose was sufficient to significantly decrease grooming behaviour. Mortality was not measured, but a previous study found that another species of leafcutter ant, Atta sexdens, had significantly increased mortality when exposed to a fungal pathogen and imidacloprid at the same concentration 10 ng/insect concentration compared to ants exposed only to the fungal pathogen (Santos et al. 2007).
Barbieri et al. (2013) exposed colonies of the Southern ant Monomorium antarcticum (native to New Zealand where the study was conducted) and the invasive Argentine ant Linepithema humile to imidacloprid in sugar water at a concentration of 1.0 μg/mL, equivalent to 1000 ng/g. Relative aggression was affected by neonicotinoid exposure, with native ants lowering their aggression to invasive ants, and conversely exposed invasive ants increasing their aggression, resulting in a lower survival probability. Brood production was not affected in the Southern ant, but exposure to neonicotinoids reduced Argentine ant brood production by 50% relative to non-exposed colonies. No effect of neonicotinoid exposure on foraging ability was detected.
Wang et al. (2015a) fed colonies of fire ants Solenopsis invicta sugar water at concentrations of 0.01, 0.05, 0.25, 0.50 and 1.00 μg/mL, equivalent to 10–1000 ng/g. The impacts on feeding, digging and foraging were quantified. Ants exposed to the 10 ng/g concentration consumed significantly more sugar water and increased digging activity. Concentrations greater than or equal to 250 ng/g significantly supressed sugar water consumption, digging and foraging behaviour.
Wang et al. (2015b) fed S. invicta newly mated queens water containing imidacloprid concentrations of 10 or 250 ng/g. Neither concentration increased queen mortality, but they did both significantly reduce queens’ brood tending ability and the length of time taken to respond to light, an indication of disturbance and colony threat. In Solenopsis species, eggs are groomed and coated with an adhesive substance that maintains moisture levels and allows for rapid transport of egg clumps. At the 250 ng/g concentration, the number of egg clumps was significantly increased (indicating low egg care and an increase in the effort needed to transport brood), suggesting that the queens had a reduced ability to groom eggs. Untended eggs become mouldy, reducing colony growth. Colonies exposed to 10 ng/g showed no difference in egg clump numbers compared to controls.
Across these ant studies, the neonicotinoid concentrations used are generally very high, in most cases far higher than expected exposure rates under field-realistic conditions (“Risk of exposure for non-target organisms from neonicotinoids applied directly to crops” and “Risk of exposure for non-target organisms from neonicotinoids persisting in the wider environment” sections). Few sublethal effects were detected at concentrations of 10 ng/g, the levels that might be reasonably expected to be encountered under field conditions.
Earthworms have similar neural pathways to insects, and earthworms are highly likely to be exposed to neonicotinoids through direct contact with soil, ingestion of organic material bound to neonicotinoids and consumption of contaminated plant material (Wang et al. 2012, “Persistence of neonicotinoids in soil” section). Reported neonicotinoid LC50s for earthworms from 13 studies range from 1500 to 25,500 ppb, with a mean of 5800 ppb and a median of 3700 ppb (see Pisa et al. 2015). Fewer studies are available that measured sublethal effects on reproduction. Negative impacts on cocoon production were measured at between 300 and 7000 ppb depending on earthworm species and neonicotinoid type.
Very little data are available for realistic neonicotinoid exposure to earthworms under field conditions. Neonicotinoid concentrations in soils can range from 2 to 50 ng/g depending on organic matter composition, application rate and other factors, although they may be much higher in immediate proximity to dressed seeds (“Persistence of neonicotinoids in soil” section). Douglas et al. (2015) detected neonicotinoids in earthworms present in thiamethoxam-treated soybean fields. Two earthworms were casually collected during soil sample collection. The two samples were found to contain total neonicotinoid concentrations of 54 and 279 ppb corresponding to ∼16 and ∼126 ng per worm. In addition to thiamethoxam and its degradates, the two earthworm samples contained imidacloprid at 25 and 23 ppb. The fields from which they were taken had not been treated with imidacloprid for at least 1 year previously, adding further to the evidence that neonicotinoids can persist in soils for over 1 year (“Persistence of neonicotinoids in soil” section). Because only live earthworms were collected and because of the small sample size, it is not clear if these are representative of typical concentrations or are an underestimate. For example, if earthworms are exposed to higher levels that cause mortality, they cannot be subsequently sampled for residue analysis.
Overall, these studies continue to increase our understanding of the negative effects of neonicotinoids on non-target organisms. In contrast to bees, most studied groups had lower sensitivity to neonicotinoids, in some cases by several orders of magnitude.
Sensitivity of aquatic invertebrates to neonicotinoids
The most comprehensive review of the acute and chronic effects of neonicotinoids on aquatic invertebrates was conducted by Morrissey et al. (2015). This followed on from and updated the reviews of Goulson (2013), Mineau and Palmer (2013) and Vijver and van den Brink (2014). Morrissey’s analysis covered 214 toxicity tests for acute and chronic exposure to imidacloprid, acetamiprid, clothianidin, dinotefuran, thiacloprid and thiamethoxam for 48 species of aquatic invertebrate species from 12 orders (Crustacea: Amphipoda (11.7% of tests), Cladocera (21.0%), Decapoda (1.9%), Isopoda (4.2%), Mysida (7.9%), Podocopida (12.6%), Insecta: Diptera (22.9%), Ephemeroptera (6.5%), Hemiptera (3.7%), Megaloptera (1.9%), Odonata (1.9%), Trichoptera (3.3%)) from peer-reviewed and government studies. Both LC50 and ED50 values were included. Acute and chronic toxicity of neonicotinoids vary greatly across aquatic invertebrates with differences of six orders of magnitude observed (Fig. 8). In general, insects were more sensitive than crustaceans; in particular, the Ephemeroptera (mayflies), Trichoptera (caddisflies) and Diptera (flies, most specifically the midges, Chironomidae) were highly sensitive.
The Cladoceran water flea D. magna was the most commonly used model organism, represented in 34 of the 214 toxicity tests (16%). Its widespread use is because of its position as a global industry standard for the majority (82%) of commercial chemicals tested (Sánchez-Bayo 2006). It shows a wide variation in sensitivity to neonicotinoids, but the mean short-term L[E]C50 is at least two to three orders of magnitude greater than for all other tested invertebrate groups (Fig. 8). This has been highlighted by several authors (e.g. Beketov and Liess 2008) who argue that given the low sensitivity of D. magna to neonicotinoids, a different model organism such as a Dipteran should be selected when conducting tests on this class of pesticide. This is illustrated by the most recent study to calculate LC50s for a range of aquatic invertebrates that was not included in Morrissey’s review. de Perre et al. (2015) found no sublethal or lethal effects of clothianidin on D. magna at concentrations of over 500 μg/L. In contrast, C. dilutus showed EC50 effects at 1.85 μg/L and LC50 effects at 2.32 μg/L, in line with previous findings (Fig. 8).
Kunce et al. (2015) also investigated the impacts of neonicotinoids on the similar C. riparius. First instar midge larvae were exposed to thiacloprid and imidacloprid at 50% of the 96-h LC50s reported in the literature, corresponding to 2.3 μg/L for thiacloprid and 2.7 μg/L for imidacloprid. Three-day-old larvae were pulse exposed to these concentrations for 1 h then transferred to clean water and allowed to develop normally. The 1-h exposure to thiacloprid significantly decreased the proportion of larvae surviving to adulthood from 94% in the control to 68%. However, imidacloprid alone and thiacloprid and imidacloprid combined had no observable effect. No difference on adult egg production levels was detected.
These recent studies in conjunction with the review of Morrissey et al. strongly support the position that insect larvae are most sensitive to neonicotinoids in aquatic environments. Morrissey et al. conclude that chronic neonicotinoid concentrations of over 0.035 μg/L or acute concentrations of over 0.200 μg/L can affect the most sensitive aquatic invertebrate species. This finding is consistent with the value suggested by Vijver and van den Brink (2014) of 0.013–0.067 μg/L for imidacloprid. A number of water quality reference values have been published by governmental regulatory bodies and independent researchers in Europe and North America (Table 8). Most of these studies are based on assessments for imidacloprid only. Values for acceptable long-term concentrations vary by three orders of magnitude from 0.0083 μg/L in the Netherlands (RIVM 2014; Smit et al. 2014) to 1.05 μg/L in the USA (US EPA 2014a). There is considerable difference in the methodologies used to calculate these reference values, with the US EPA value likely to have been strongly based on results from D. magna, a species known to have relatively low sensitivity to neonicotinoids (Morrissey et al. 2015).
Table 8 Summary of published ecological quality reference values for neonicotinoids (imidacloprid except this review) in freshwater environments against which average (chronic or long-term) or maximum (acute or peak) exposure concentrations are to be compared
Current levels of neonicotinoids in aquatic habitats regularly exceed this threshold, as discussed in “Levels of neonicotinoid contamination found in waterbodies” section. Combining the review of Morrissey et al. (2015) with recent publications, a total of 65.3% of studies (17/26) report average neonicotinoid concentrations over the 0.035 μg/L chronic threshold and 73.5% of studies (25/34) report peak concentrations over the 0.200 μg/L acute threshold. The number of countries that have been studied and their widespread distribution (Australia, Brazil, Canada, China, Hungary, Japan, the Netherlands, Sweden, Switzerland, the USA and Vietnam) indicates the widespread contamination of watercourses of all kinds with levels of neonicotinoids known to be harmful to sensitive aquatic invertebrates. This is now a chronic global problem, likely to be impacting significantly on aquatic insect abundance and on food availability for their predators, including fish, birds and amphibians.
Sensitivity of birds and bats to neonicotinoids
Gibbons et al. (2015) reviewed the direct and indirect effects of neonicotinoids and fipronil on vertebrate wildlife including mammals, fish, birds, amphibians and reptiles. LD50 values for imidacloprid, clothianidin and fipronil are available for 11 species of bird (Table 9). There is considerable variation in the lethality of these compounds to birds, both between bird species and pesticide type. Using US EPA (2012) classifications for toxicity (see legend for Table 9), imidacloprid ranged from moderately toxic to highly toxic, clothianidin from practically non-toxic to moderately toxic and fipronil from practically non-toxic to highly toxic. Many of these studied bird species are granivorous and can be expected to feed on sown seeds shortly after the sowing period. Theoretical levels of seed consumption necessary to cause mortality were calculated by Goulson (2013); see “Risk from non-flowering crops and cropping stages prior to flowering” section.
Table 9 Single (acute) dose LD50 for bird species (ng/g) for imidacloprid, clothianidin and fipronil
In addition to lethal effects, several studies have identified sublethal effects of neonicotinoid ingestion on birds (Table 10). House sparrows can become uncoordinated and unable to fly, and studies of Japanese quail and red-legged partridges have reported DNA breakages and a reduced immune response, respectively. Many of these sublethal effects occur at lower concentrations than the lethal dose. A single oral dose of 41,000,000 ng/g of imidacloprid will cause mortality in house sparrows; a substantially lower dose (6000,000 ng/g) can induce uncoordinated behaviour and an inability to fly (Cox 2001). Whilst imidacloprid is highly toxic to Japanese quail, with an LD50 of 31,000,000 ng/g, chronic daily doses of 1000,000 ng/g/day can lead to testicular anomalies, DNA damage in males and reductions in embryo size when those males are mated with control females (Tokumoto et al. 2013).
Table 10 Other studies of the direct effects of imidacloprid, clothianidin and fipronil on birds
In addition to the studies reviewed by Gibbons et al., one additional study is available that assessed the impact of neonicotinoid ingestion on birds. Lopez-Anita et al. (2015) fed red-legged partridge Alectoris rufa imidacloprid-treated wheat seeds for a period of 25 days in the autumn and an additional period of 10 days in the spring, matching the pattern of cereal cropping in Spain. One treatment contained seeds treated at the recommended dosage rate and the second at 20% of the recommended rate, to mimic a diet composed of 20% of treated seeds. Treated seeds contained concentrations of imidacloprid of 140,000–700,000 ng/g at the two dose rates. As the 400 g partridges used in this study consume around 25 g of seeds a day, a daily ingestion of 8800 and 44,000 ng/g/day was expected.
Imidacloprid at the highest dose killed all adult partridges in 21 days, with first deaths occurring on day 3. Mortality in the low dose and control groups was significantly lower at 18.7 and 15.6% respectively. As all partridges in the high dose died, effects on reproductive output were only measured in the low dose treatment. Compared to controls, low dose females laid significantly smaller clutches, and the time to first egg laying was also significantly increased. There was no difference in egg size, shell thickness, fertile egg rate and hatching rate. There was no detectable impact on chick survival, chick growth or sex ratio between these two groups. These results are in line with previous findings for lethal (Table 9) and sublethal (Table 10) effects of neonicotinoid consumption by birds. Whilst LD50s vary across two orders of magnitude from 11,300 to >2000,000 ng/g, sublethal effects are seen across a more consistent range of doses over one order of magnitude between 1000 and 53,000 ng/g. The greatest outstanding issue is that no data exist that quantify the actual exposure rate to granivorous birds from neonicotinoid-treated seeds. As such, it is difficult to judge whether these clearly demonstrated lethal and sublethal effects are manifested in wild bird populations in the field.
In addition to sublethal and lethal effects potentially caused by the ingestion of neonicotinoids from treated seeds, bird populations may also be affected by a reduction in invertebrate prey. Hallmann et al. (2014) used bird population data from the Dutch Common Breeding Bird Monitoring Scheme, a standardised recording scheme that has been running in the Netherlands since 1984. Surface water quality measurements are also regularly collected across the Netherlands, including data on imidacloprid levels. Hallmann et al. compared surface water imidacloprid levels between 2003 and 2009 with bird population trends for 15 farmland bird species that are insectivorous at least during the breeding season to assess the hypothesis that neonicotinoids may cause bird population declines through a reduction in invertebrate food availability. The average intrinsic rate of increase in local farmland bird populations was significantly negatively affected by the concentration of imidacloprid. At the individual level, 14 of the 15 bird species showed a negative response to imidacloprid concentrations, with 6 out of 15 showing a significant negative response. As previously discussed in “Sensitivity of butterflies and moths to neonicotinoids” section, it is difficult to disentangle the effects of neonicotinoids from the effects of general agricultural intensification. Hallmann et al. attempt to control for proxy measures of intensification including changes in land use area, areas of cropped land and fertiliser input, but imidacloprid levels remained a significant negative predictor.
The only available study that has quantified changes in invertebrate prey availability after neonicotinoid treatment and concurrent changes in the bird community was conducted in the USA. Falcone and DeWald (2010) measured invertebrates in eastern hemlock Tsuga canadensis forests in Tennessee after trees have been treated with imidacloprid to control hemlock woolly adelgid Adelges tsugae. The imidacloprid treatment had a significantly negative effect on non-target Hemiptera and larval Lepidoptera. However, there was no corresponding decline in insectivorous bird density between treatments. Direct comparison between this study and the findings of Hallmann et al. 2014 are difficult due to the very different ecological conditions. It is likely sufficient untreated areas existed in hemlock forests for insectivorous birds to find sufficient forage. In the Netherlands, one of the most agriculturally intensified regions in the world, unaffected semi-natural habitat is scarce and a reduction in prey availability caused by neonicotinoid application would have a more severe impact.
No studies are available that measure the effect of neonicotinoids on bats and bat populations. A link between neonicotinoid use and declining farmland butterfly populations has been suggested (Gilburn et al. 2015; Forister et al. 2016), and given the ecological similarity between butterflies and moths, a similar trend may be ongoing, though this has not yet been investigated. Many bat species feed on moths, so a reduction in the moth population is likely to impact bat populations through a reduction in food availability. Mason et al. (2014) link neonicotinoid use with an increase in the frequency of bat diseases such as White Nose Syndrome (caused by the fungus Geomyces destructans) in both the USA and Europe. They hypothesise that consumption of neonicotinoid residues in insect prey weakens the immune system of bats. However, no evidence is presented demonstrating the presence of neonicotinoid residues in moths or bats or a passage across these trophic levels or that exposure to neonicotinoids weaken the immune system of bats, resulting in increased rates of fungal infection. The position of Mason et al. must currently be considered unsupported.
Synergistic effects of additional pesticides with neonicotinoids
The EFSA (2013a, b, c) risk assessments for clothianidin, imidacloprid and thiamethoxam considered these pesticides and their impacts on honeybees individually. In the field, multiple neonicotinoids, other insecticides and other pesticides such as herbicides and fungicides are commonly applied to a single crop. Bees are frequently exposed to complex mixtures of pesticides, with 19 detected in trap-caught bees from an agricultural region of Colorado (Hladik et al. 2016). It is possible that combinations of neonicotinoids and other pesticides may have antagonistic (become less effective), additive (equivalent to adding together existing effectiveness) or synergistic (multiplicative) effects. Morrissey et al. (2015) briefly listed known examples of synergistic effects between neonicotinoids and other pesticides. Several examples have been demonstrated by pesticide companies themselves. For example, Bayer demonstrated that the combination of clothianidin and the fungicide trifloxystrobin resulted in a 150-fold increase in kill rate to Phaedon leaf beetle larvae over clothianidin alone (Wachendorff-Neumann et al. 2012). Bayer scientists also demonstrated that treatments of 8000 ppb of thiacloprid and 8000 ppb of clothianidin resulted in aphid population kill rates of 25 and 0% after 6 days. Combining the two increased the kill rate to 98% (Andersch et al. 2010). Specifically for honeybees, Iwasa et al. (2004) demonstrated that the combination of thiacloprid with the fungicide propiconazole increased the toxicity of the mixture several hundred fold. Whilst synergies have been demonstrated, few environmental risk assessments have been made for neonicotinoids in combination with other pesticides.
Since 2013, a number of studies have investigated possible synergistic effects in neonicotinoids. Several have focused on the interaction between neonicotinoids and ergosterol biosynthesis inhibitor (EBI) fungicides (which include propiconazole) and their impact on bees. Biddinger et al. (2013) studied the interaction between the contact toxicity of acetamiprid, imidacloprid and the fungicide fenbuconazole, a substance virtually non-toxic to bees (except at extremely high concentrations), using A. mellifera and Japanese orchard bees Osmia cornifrons. These pesticides are commonly found together in tank mixes used in orchards. The doses ranged from 1.38 to 60 μg/bee 1:1 acetamiprid plus fenbuconazole mixture and 0.86 to 983 μg/bee 2:1 imidacloprid plus fenbuconazole mixture. At LD50, the acetamiprid and fenbuconazole mixture was ∼5 times more toxic than acetamiprid alone for A. mellifera and ∼2 times more toxic than acetamiprid for O. cornifrons. However, these doses are exceptionally high, for example the 0.86 μg/bee imidacloprid/fenbuconazole mixture is equivalent to 567.6 ng/bee, with the A. mellifera contact toxicity to imidacloprid LD50 calculated as 81 ng/bee (“Sensitivity of bumblebees and solitary bees to neonicotinoids” section). Unsurprisingly, this dose killed 85% of honeybees in this treatment. At unrealistically high concentrations, it is not clear how informative these results are.
Thompson et al. (2014) investigated synergies between several EBI fungicides (flusilazole, propiconazole, myclobutanil and tebuconazole) and a range of neonicotinoids (clothianidin, thiacloprid, imidacloprid and thiamethoxam) on A. mellifera. Individual pesticides and mixtures of one neonicotinoid and one fungicide were administered through both contact and ingestion at a range of concentrations sufficient to increase mortality, and bees were observed for a 96-h period. LD50s were calculated after 48 h as mortality did not significantly increase after this point. Single neonicotinoid and fungicide doses showed similar toxicity to previous published results, with no individual fungicide causing toxic effects even at concentrations of 22.4 μg/bee.
For neonicotinoid/fungicide mixtures, neonicotinoids were applied at calculated LD50s, in the region of 0.035–0.124 μg/bee for clothianidin, imidacloprid and thiamethoxam and 122.4 μg/bee for thiacloprid (cyano-substituted neonicotinoids having lower toxicity to bees, “Direct lethality of neonicotinoids to adult wild bees” section). Fungicides were applied at doses of between 0.161 and 0.447 μg/bee depending on the particular compound. These values were calculated as realistic worst-case exposures based on approved application rates for UK crops. For these mixtures, a synergy ratio was calculated where the LD50 of the neonicotinoid was divided by the LD50 of the neonicotinoid plus fungicide mixture. Consequently, a value of over 1 indicates that the mixture was more toxic and a value under 1 indicates that the mixture was less toxic. Combinations of fungicides with thiacloprid and clothianidin showed negligible synergy for contact toxicity, with an average synergism ratio of 0.30 and 1.07 respectively. Imidacloprid and thiamethoxam were higher at 1.53 and 2.02. For oral toxicity, thiacloprid and imidacloprid showed low synergy at 0.60 and 0.48 whereas clothianidin and thiamethoxam were higher at 1.52 and 1.31 respectively. Only two combinations showed significant synergy, for a contact dose of tebuconazole and thiamethoxam with a synergy of 2.59 and for an oral dose of clothianidin and tebuconazole at a synergy of 1.90.
Sgolastra et al. (2016) investigated the interaction between clothianidin and the fungicide propiconazole in three bee species, A. mellifera, B. terrestris and O. bicornis. Each species was administered a LD10 dose of clothianidin (0.86, 1.87 and 0.66 ng/bee respectively; see “Direct lethality of neonicotinoids to adult wild bees” section for more details), a non-lethal dose of propiconazole (7 μg/bee) and a combination of the two treatments. Bees were then observed for a 96-h period and mortality quantified. Some synergistic effects were seen. In A. mellifera, mortality was significantly higher for the combined dose in the first two time periods (4 and 24 h). Mortality in B. terrestris for the combined dose was only significantly higher in the first time period, after 4 h. However, in O. bicornis, exposure to the combination of clothianidin and propiconazole resulted in significantly higher mortality at all time points.
Spurgeon et al. (2016) conducted similar experiments to Sgolastra et al., investigating the effect of a combination of clothianidin and propiconazole on A. mellifera, B. terrestris and O. bicornis. In order to calculate an LD50, clothianidin concentrations were varied and propiconazole concentrations were held at zero, a low dose and a high dose. The low dose was taken from the EFSA Panel on Plant Protection Products (2012) reported environmental concentrations, and the high dose was 10 times the low dose to represent a plausible worst case scenario, but it is not clear what these values actually are. Mortality was quantified over 48, 96 and 240 h. For A. mellifera, clothianidin LC50s with and without propiconazole were always within a factor of 2, with no clear negative trend at higher propiconazole concentrations. For B. terrestris, clothianidin LC50s with propiconazole were between 1.5 to 2 fold lower. For O. bicornis, clothianidin LC50s with propiconazole was up to 2-fold lower with a negative trend as propiconazole concentrations increased. Spurgeon et al. concluded that the clothianidin and propiconazole combination had no to slight synergy for A. mellifera and slight to moderate synergy for B. terrestris and O. bicornis.
In an additional trial, Thompson et al. (2014) demonstrated that the dose of fungicide applied is a key factor determining neonicotinoid toxicity using propiconazole and thiamethoxam mixtures (Table 11). The authors argue that their low rates of significant synergies between neonicotinoids and fungicides was because of their lower, more field-realistic fungicide doses of 0.161–0.447 μg/bee compared to 10 μg/bee used by Iwasa et al. (2004), an early study demonstrating this interaction. The values of 0.161–0.447 μg/bee were calculated as realistic worst-case exposures based on approved application rates for UK crops. However, data are lacking that demonstrate true field-realistic exposure rates to fungicides for free flying bees. Whilst studies such as Sgolastra et al. (2016) show a clear synergistic effect between fungicides and neonicotinoids on O. bicornis, the dose of fungicide used is more than an order of magnitude greater than that used by Thompson et al. Bees are consistently exposed to fungicides with 40 types found in honeybee pollen, wax and nectar (Sánchex-Bayo and Goka 2014). Pollen collected by bumblebees and stored in their nests has also been found to contain fungicides at average concentrations between 0.15 and 25 ppb (EBI fungicides 0.15–17 ppb; David et al. 2016). However, almost nothing is known about how concentrations present in bee-collected material translate into acute or chronic exposure to bees. It is currently not known what fungicide doses represent a realistic situation that bees are likely to encounter in the wild, though models such as Bee-REX are attempting to bridge this gap (US EPA Agency 2014b).
Table 11 Comparison of the ratio of propiconazole to the doses of thiamethoxam and the resultant LD50 in the contact and oral studies
In addition to work on bees, Kunce et al. (2015) investigated the impact of 1-h pulse exposure of imidacloprid and thiamethoxam and two pyrethroids, deltamethrin and esfenvalerate, in single, pairwise and combined doses on the development of the aquatic midge C. riparius (see “Sensitivity of aquatic invertebrates to neonicotinoids” section for more methodological and concentration details). Most pesticide treatments reduced the survival of the larvae, but the deleterious effects did not appear to be synergistically amplified by a combination of pesticides. Kunce et al. conclude that at the low doses and period of exposure used, the risk of synergistic or additive effects is very low.
Overall, these studies support the position that neonicotinoids can act synergistically with fungicides, increasing their lethality to bees. However, the dose rate of both neonicotinoids and fungicides, time of exposure, neonicotinoid and fungicide chemical class and length of time after exposure are all important explanatory factors affecting this relationship. The concentration of fungicide used in laboratory studies appears to be the most important factor determining synergistic lethality. Fungicides are regularly sprayed during the period when flowering crops are in bloom under the assumption that these compounds are safe for bees, but this work demonstrates that their toxicity is contingent on other chemical management choices at a site. Studies to date have only examined pairwise interactions between pesticides. It is clear that bees and other non-target organisms inhabiting farmland are routinely exposed to far more complex mixtures of pesticides than any experimental protocol has yet attempted to examine. For example, honeybee and bumblebee food stores commonly contain 10 or more pesticides (e.g. David et al. 2016). A major challenge for scientists and regulators is to attempt to understand how chronic exposure to complex mixtures of neonicotinoids and other chemicals affects wildlife.