A recent American study - Mogren and Lundgren 2016 - it found that the pollen and nectar in flowers in pollinator strips adjacent to crops were so polluted with neonicotinoids that they caused harm to Honeybees and that the concentrations of neonics in bee bread were eye wateringly high, well above the levels that could stop wild bees from reproducing successfully (e.g. Whitehorn et al. 2012). While Mogren and Lundgren 2016 report that exposure of bees to neonics was much higher close to neonic treated crops, even wildflower strips 150 metres away from treated crops caused significant exposure to the Honeybees. The most important caveat is, of course, that this study was in the US where treatment rates and frequency may be more intense than in the EU.

Planted crop margins do provide lots of great resources for bumblebees, butterflies and other insects (Pywell et al. 2011, Carvell et al. 2005, Carvell et al. 2006, Blake et al. 2011, Thomas and Marshall 1999, Haaland et al. 2011). Although caution is required in drawing conclusions about the population effects of flower resources from foraging activity data, for instance Holzschuh et al. 2016 found that despite certain pollinators feeding on mass flowering crops, it did not follow that at a landscape scale their populations benefited.

The risk to field margin ecology from pesticides used on adjacent crops has long been recognised, in 2002 Marshall and Moonen said “Their close proximity to agricultural operations renders them susceptible to disturbance, particularly from pesticide drift and eutrophication”. Even the pesticide industry’s own report on crop field margins (Hackett and Lawrence 2014) acknowledges that pesticide contamination could harm life in the field margins. “Field margins can also separate the cropped area from hedgerows or other off-crop features (including other cropped areas) thus reducing levels of spray drift of pesticides. However, to reduce spray drift to a hedgerow the field margin intercepts the pesticide and may be impacted. The level of impact caused by spray drift will then depend on whether a species occupies the field margin or hedgerow alone.” and it is known that water transfer can move pesticides up to 20 meters across field margins “the CORPEN9 review, and the guidance produced, which recommends buffer width of between 10 m to 20 m for 70 to 80 % reduction efficiency of pesticides.”

Despite the clear risk to crop margin habitat strips from pesticide sprays there not many papers quantifying the impacts of insecticide sprays on field margins, and none on bees that we can find! However, Bundschuh et al. 2012 found that only field margins wider than nine metres supported grasshoppers and put this down to pesticide drift; Langhof et al. 2009 estimated that pesticide drift three metres into a margin would cause <30% [!] mortality of a parasitic wasp and ≤52% mortality of seven spot ladybirds; and Hahn et al. 2015 found that pyrethroid insecticide sprays significantly reduced the abundance of caterpillars and moths in crop field margins.

The paucity of evidence relating to insecticide sprays and pollinators in field margins is unfortunate considering that pyrethroids in particular are known to impact on beneficial invertebrates (Ewald et al. 2016) and their use is very high – almost universal on arable land – and increasing (Underwood and Mole 2016). We really should by now have properly ascertained the risk to pollinators from sprays drift onto field margins, thankfully there is more information about the movement of neonicotinoid seed treatments into margins and the levels of exposure that create lethal and sub-lethal harm to pollinators.

So do we know if neonic levels in EU planted crop margins were so high that their toxicity out-weighed the benefits to pollinators from the pollen and nectar supplied?

The levels of neonics found in the USA field margin study are high, but high levels have also been found in field margins in the EU. In particular Botias et al. 2015 (and Botias et al. 2016) found higher levels of neonics in wildflowers in margins than in adjacent Oilseed rape crops, including one plant with a very high concentration indeed, the study also found that 97% of Honeybee neonic exposure was via wildflowers, with only 3% via the crops.

The findings of higher levels of neonics in wildflowers than in the adjacent crop is not unusual and was also found by Stewart et al. 2014 and Rundlof et al. 2015. These studies focus on concentrations in wildflowers which may not be representative of the concentrations that occur in plants in planted crop margins, however Botias et al. 2015 did find high concentrations in Vicia and Trifolium and Stewart et al. 2014 also included these genera commonly in their analysis; so plants commonly grown in pollinator strips are definitely capable of transmitting harmful levels of neonics to pollinators. Different types of plants take up, concentrate or accumulate neonics differently. We know for instance that crops show different preferential absorption of neonicotinoids and take them up to different extents Sur and Stork 2003, so some wildflowers may be more prone to concentrating neonicotinoids than some crops.

I have had a previous discussion with Christina Botias about this possible phenomenon and she said “We didn´t find a very clear trend of a specific plant or plant types (woody vs herbaceous, perennial vs annual) to be more likely contaminated. We detected clothianidin at significantly higher concentrations in annual plants vs. perennials, but then imidacloprid was present at higher concentrations in perennials. Also imidacloprid was at higher levels in herbaceous vs wood plants.”

Of course there are lots of other studies that detected high levels of neonics in pollen and nectar in field margin plants, including Greatti et al. 2006, Krupke et al. 2012, Pettis et al. 2013 (data), David et al. 2016 and Mortl et al. 2018.

In the absence of a tenable theory why they would be less polluted, in my view there is no reason not to suppose that plants in crop margins would have been just as polluted as other wildflowers growing in similarly highly exposed situations.

It may be thought that most of the contamination of wildflowers adjacent to crops was via the movement of soil water, but dust probably also played a key role in exposing plants and pollinators near crops to dangerous levels of neonics. Dust emitted during the planting of neonic treated seeds contained very high concentration of neonicotinoids Girolami et al. 2011, the dust landed directly on wildflower leaves where it could be readily absorbed, a leaf presents a much larger and more permeable surface area than a seed.

Neonicotinoids appear to concentrate in the soil surface, and when the field is bare toxic dust can blow between fields Limay-Rios et al. 2015, affecting large areas (Krupke et al. 2017) and indeed for distances over 250 m Forero et al 2017 – so plants growing in margins when field is bare were/are vulnerable to additional contamination from neonics concentrated in such dust.

We know from Woodcock et al. 2016 that 40% of the wild bees they studied had disappeared from at least 10% of their UK distribution as a direct result of neonicotinoid use. Neonics were not simply reducing wild bee abundance; the effect was so strong that they also caused bee species to entirely disappear from large parts of their former range. Figure 2 in Woodcock et al. 2016 shows that the negative shift on 2a due to neonicotinoid exposure for Oilseed rape (OSR) foraging bees is considerably greater than the positive shift for OSR-feeders associated with increased area of OSR in 2b and the negative shift in 2a for non-OSR-feeding bees is additive to the negative shift in 2b. Therefore, there were marked negative impacts from OSR+neonics on both groups of bees. But, how much of the neonicotinoid exposure arose from feeding on Oilseed rape, as assumed in Woodcock et al. 2016, and how much arose from contaminated wildflowers? We do not have a conclusive answer to this, while Botias et al. 2015 found that 97% of the exposure was from wild flowers, other studies recorded higher proportions of pollen and nectar being gathered from the crop. On the other hand, many of the studies only focussed on the exposure of the bees while Oilseed rape was in flower and hence ignored exposure from adjacent wildflowers during the rest of the year. We will probably never know the balance of contamination exactly, but most recent research suggests that contamination from non-crop flowers was in the same ball park as crop mediated contamination (Botias et al. 2015 (10% of pollen from OSR), Tsvetkov et al. 2017, Garbuzov et al. 2015, Long and Krupke 2016). David et al. 2016 is particularly helpful here, they found that pollen from OSR contained 166 ng/g of pesticide, compared with from 78 and 25 ng/g for wildflowers sampled from OSR and Winter wheat margins respectively. However, in Honeybee collected pollen they found 17 ng/g during OSR flowering and 2.6 ng/g after. While there is little doubt that contamination levels during the 3 or 4 weeks of OSR flowering was higher than the contamination at other times, contamination from wildflowers occurred over six to eight times the time period, so I suspect that in in practice the annual dose from each source – crop and wildflowers - was often similar.

Another study gives us some additional insight; Tsvetkov et al. 2017 studied the exposure of bees in a system in Canada where there were no crops regularly visited by bees, so almost all the exposure must have come from wildflowers. They found levels of contamination that caused significant harm to Honeybee health, showing again that neonics do not need a crop vector to cause harm to pollinators.

I think it is an unavoidable conclusion that neonicotinoid contamination from crop margins has the potential to harm bees, but is there evidence that the volume of nectar and pollen produced in planted pollen and nectar strips countered this harm to pollinators? Despite their pollution with neonicotinoids, did flower rich crop margins continue to provide a net benefit to wild pollinator populations?

Direct evidence showing benefits of particular flower-rich habitats to pollinator populations is rare due to the difficulties of measuring this. It is one thing to record if pollinators are using a pollinator margin, but it is a much harder task to examine whether this is having a population level positive effect.

Key studies such as Wood et al. 2016 – that showed that planted crop margins provide food resources for bumblebees and Honeybees, but are a lot less helpful to solitary bees – rely on counting visits by bees and analysing their pollen loads. Unfortunately this is not data that helps us to prove that the margins are having a positive effect at a population level.

One of the few studies to look at actual populations of bees at a landscape scale during the height of the neonicotinoid contamination period was the CEH study on the Hillesden Estate (Carvell et al. 2017 and Redhead et al. 2016). This study used advanced molecular techniques coupled with detailed surveys of floral resources to show a greater residence time of bumblebee family lineages (and therefore benefits to the population) in arable landscapes where flower-rich habitats had been sown compared with arable landscapes depauperate in floral resources. This study was conducted across the 1000ha Hillesden experimental farm in 2011-2012, while the farm was using neonicotinoid seed dressings on its Oilseed rape and wheat. The study was also coincident with the very widespread use of neonicotinoids in these crops elsewhere in the landscape. So we can infer that despite the use of neonicotinoids, the provision of floral resources appeared to benefit bumblebee populations. At Hillesden the majority of the floral habitats were perennial wildflowers (by area) but there were also patches of annual wild bird seed. However, what we do not know from this study is whether the benefits of floral resources would have been greater if neonicotinoids had not been used on the farm and wider landscape, nor is it possible to split apart the comparative effects of restored meadows, versus arable margins (Claire Carvell pers. com.), because the data is lumped, one habitats could be providing a benefit and the other harm and still give a net positive result – particularly as planted crop margins were a minor component of the provided floral resources (only 2% of total area).

Recent evidence Powney et al 2019 gives us some improved confidence that even at the height of the use of neonics the creation of flower rich field margins may have continued to provide significant benefit to some pollinator species. The data shows that, in contrast to other bees and hoverflies, eusocial species (primarily bumblebees and honey bees) maintained their distribution between 2002-2013. These species have a big overlap with those that we know benefit from flower rich habitats on arable margins. So something appears to have been sustaining their populations while other species were hit hard by neonics, and it may well have been field margins in agri-env schemes that enabled this to happen. Although we should remember that only <0.2% of England is in flower rich margins and other factors could be responsible for eusocial species doing better than solitary bees.

Neonicotinoids have now been comparatively well studied, so we have more knowledge about their persistence, dispersal and action in the environment than we do about most pesticides. However, the three neonics that are subject to the 2018 ban are not the only chemicals we need to consider when thinking about the potential for pesticides to harm bees. In the USA a link has been found between Chlorothalonil and bumblebee declines McArt et al. 2017. From a domestic perspective, although it is due to be phased out in 2020, in 2016 Chlorothalonil was the most widely-used individual active substance and in terms of weight applied, the principal formulation used in the UK Fera 2018. In addition there are other persistent insecticides now coming into use such as Cyantraniliprole and even new neonicotinoids such as Sulfoxaflor are awaiting approval. To date the 2013 EFSA bee risk assessment process is being blocked by the EU Member States, in the absence of any new measures to prevent future pesticides from causing a recurrence of the harm caused to bees and pollinators by neonics we have to assume that this will recur, and may even be worse next time.

So to summarise the evidence; neonics in the landscape harmed bee populations (and probably also populations of birds Hallmann et al. 2014, butterflies Gilburn et al. 2015 and other animals Douglas et al. 2015) and a significant proportion of this harm arrived through flowers growing adjacent to and within 150 metres of treated crops. In some cases harm was observed in landscapes where none of the contamination would have come through the pollen and nectar of a crop. There is good evidence that before neonics the pollen and nectar boost provided by planted crop margins assisted bumblebee populations and Honeybees. There is evidence that these eusocial bees were less impacted by neonicotinoids than other bees, but it remains unclear if the harm caused by the accumulating levels of neonicotinoids entering these and many other pollinator species through the crop margins negated the benefit that was previously provided. It is not clear if the most polluted crop margins were causing net harm, although it seems likely, or if there were sufficient less polluted crop margins that counteracted this and ensure that such agri-environment schemes therefore continued to provide a net benefit to pollinator populations, although this seems likely for eusocial bees. Of course bees are not the only beneficial or conservation significant species of invertebrate that may benefit from well managed field margins, but we have practically no information about the impacts of neonicotinoids in field margins on hoverflies, ground beetles, lacewings, ladybirds, butterflies or moths.

It is tempting to breathe a sigh of relief and think that thankfully now the three most persistent neonicotinoids have been banned crop margins will again provide a clear benefit. But what of Chlorothalonil, Cyantraniliprole, Sulfoxaflor and pesticides still in the pipeline – could they cause similar or worse harm? – no-one has done the science so we do not know. In addition pyrethroid spray drift may also impact on bees using field margins (as it does with moths), but this is yet to be quantified. Until the pesticide testing regime is improved to include independently run testing of the persistence, fate and impacts of such chemicals on wild bees – before they are approved for use – planted crop margins remain vulnerable to becoming part of a poison delivery mechanism to bees and other pollinators.

Society wants public investment to restore wildflowers to the countryside but we must be mindful of the potential for pesticide contamination to reduce the value of this investment. So we do not put all our eggs in one basket, and as we cannot yet be confident that plants growing close to crops will not be toxic to pollinators, it may be wise to also target resources towards restoring large areas of flower-rich habitats that are capable of supplying food resources to solitary, especially oligolectic, bees; that can also provide nesting habitat and undisturbed soil faunas; and that have central areas that provide some refuge from the higher levels of pesticide contamination found in field margins.

The Call for Knowledge on the impacts of pesticide and fertiliser use in farmland on the effectiveness of adjacent pollinator conservation measures such as flower strips and hedgerows is still very much open to any other input in terms of studies, practical experiences etc, so please feel free to add to the KNOCK Forum and to let any of your colleagues and friends to add to the Call!

Semi-natural habitats benefit pollinators that can support pollination services to wild plants and crops. There is some published evidence of widespread pesticide contamination of non-crop flowers and fewer studies indicating that landscape-scale semi-natural habitat may mitigate effects of conventional intensive agriculture (including pesticide use) on pollinators to a certain degree. Below I detail some key points:
• Land-use change, conventional agricultural management and pesticide use represent a major risk to pollinators and pollination, but agricultural management more sympathetic to beneficial biodiversity can be part of the solution (IPBES, 2016; Kovács-Hostyánszki et al., 2017; Potts et al., 2016).
• Conclusions of a global meta-analysis (Kennedy et al., 2013) on local and landscape effects on wild bee pollinators were that farm-scale simplification of fields (monoculture) increases the importance of the quantity and diversity of semi-natural in the surrounding landscape. Conversely, field diversification lowers this reliance on landscape quality for bees.
• There is evidence of positive relationships between native bee richness, abundance and flower visitation and landscape-scale semi-natural habitat and negative relationships with agricultural management intensity (including pesticide use or proxies thereof) (Kennedy et al., 2013; Nicholson et al., 2017). There are links between semi-natural habitat, ecological restoration, pollinator visitation and diversity and pollination of crops and wild plants (Garibaldi et al., 2016; IPBES, 2016; Kovács-Hostyánszki et al., 2017; Pywell et al., 2015).
• There is also evidence that organic farms (with low or no pesticide use) tend to support greater local numbers and richness of foraging insect pollinators, and some evidence that it can benefit pollination, although this effect tends to be reduced in already diverse, heterogeneous landscapes (IPBES, 2016; Kennedy et al., 2013).
• On-farm semi-natural habitats, such as hedgerows and sown flower margins, provide food and nesting resources for insects, including pollinators and natural pest control agents, increasing their activity, and with emerging evidence of population benefits (Carvell et al., 2017; Haenke et al., 2014; Jha and Kremen, 2013; Kremen et al., 2018; Ponisio et al., 2016).
• A limited number of studies show that increasing the proportion of natural habitat in the surrounding landscape can buffer the effects of farm pesticide use on wild bee abundance and species richness. Park et al. (2015 observed pesticide effects on a wild bee community visiting an apple (Malus domestica) orchard were buffered by increasing proportion of natural habitat in the surrounding landscape. Bee communities on more intensive farms in areas with little semi-natural habitat in the surrounding local landscape were less abundant and diverse with a corresponding lowering of visitation to crop flowers (blueberry) compared to areas with abundant natural cover in the landscape (Nicholson et al., 2017).
• The interaction between pesticide load and semi-natural habitat is likely to produce complex responses according to taxonomic identity of the organism. For instance, wild bees, true bugs and ground beetles had stronger responses (community homogenisation) to habitat fragmentation at high pesticide loads, whereas for plants and spiders landscape structure was less influential at high pesticide levels (Dormann et al., 2007).
• Non-cultivated plants in agricultural landscapes are a major source of floral resources for bees (Requier et al., 2014). Contamination of pollen from these non-crop sources by multiple pesticide residues appears to be widespread and common (Botías et al., 2015; Long and Krupke, 2016; McArt et al., 2017). This suggests a potential pathway of pesticide exposure to pollinators from spillover or soil contamination of adjacent non-crop habitat (perennial or established annually).
• There appears, however, to be a dearth of empirical knowledge about whether an interplay between pesticide use in fields and the presence of adjacent field margin habitats affects pollinator diversity, abundance, species interactions and plant pollination. We do not know if providing ecological infrastructure on farms can mitigate the effects of pesticide exposure in fields. Moreover, how much semi-natural habitat is required to achieve this, or what level of floral resource diversity in space or time can lower the risk from foraging on pesticide treated crops? Equally, the extent that pesticide-use in conventionally managed fields lowers the efficacy of on-farm semi-natural habitats or ecological restoration measures (hedgerow, sown flower margins) that aim to support populations or diversity of pollinators is not established.
• One way for EKLIPSE to address the issue of whether there is an effect on pollinators and pollination from an interaction between pesticide use and ecological infrastructure would be to convene a small expert group. They could rapidly assess the literature that can provide insight to this question and qualitatively rate the likelihood of harm, form hypotheses to be tested, and scope research themes and/or approaches that are relevant to informing policymaking.

References:
Botías C, David A, Horwood J, Abdul-Sada A, Nicholls E, Hill E, et al. Neonicotinoid Residues in Wildflowers, a Potential Route of Chronic Exposure for Bees. Environmental Science & Technology 2015; 49: 12731-12740.
Carvell C, Bourke AFG, Dreier S, Freeman SN, Hulmes S, Jordan WC, et al. Bumblebee family lineage survival is enhanced in high-quality landscapes. Nature 2017; 543: 547-549.
Dormann CF, Schweiger O, Augenstein I, Bailey D, Billeter R, De Blust G, et al. Effects of landscape structure and land-use intensity on similarity of plant and animal communities. Global Ecology and Biogeography 2007; 16: 774-787.
Garibaldi LA, Carvalheiro LG, Vaissière BE, Gemmill-Herren B, Hipólito J, Freitas BM, et al. Mutually beneficial pollinator diversity and crop yield outcomes in small and large farms. Science 2016; 351: 388-391.
Haenke S, Kovács-Hostyánszki A, Fründ J, Batáry P, Jauker B, Tscharntke T, et al. Landscape configuration of crops and hedgerows drives local syrphid fly abundance. Journal of Applied Ecology 2014; 51: 505-513.
IPBES. The assessment report of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services on pollinators, pollination and food production. S.G. Potts, V. L. Imperatriz-Fonseca, and H. T. Ngo, (eds). Secretariat of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services, Bonn, Germany, 2016.
Jha S, Kremen C. Resource diversity and landscape-level homogeneity drive native bee foraging. PNAS 2013; 110: 555-558.
Kennedy CM, Lonsdorf E, Neel MC, Williams NM, Ricketts TH, Winfree R, et al. A global quantitative synthesis of local and landscape effects on wild bee pollinators in agroecosystems. Ecology Letters 2013; 16: 584-599.
Kovács-Hostyánszki A, Espíndola A, Vanbergen AJ, Settele J, Kremen C, Dicks LV. Ecological intensification to mitigate impacts of conventional intensive land use on pollinators and pollination. Ecology Letters 2017; 20: 673-689.
Kremen C, M'Gonigle LK, Ponisio LC. Pollinator community assembly tracks changes in floral resources as restored hedgerows mature in agricultural landscapes. Frontiers in Ecology and Evolution 2018; 6.
Long EY, Krupke CH. Non-cultivated plants present a season-long route of pesticide exposure for honey bees. Nature Communications 2016; 7.
McArt SH, Fersch AA, Milano NJ, Truitt LL, Boroczky K. High pesticide risk to honey bees despite low focal crop pollen collection during pollination of a mass blooming crop. Scientific Reports 2017; 7.
Nicholson CC, Koh I, Richardson LL, Beauchemin A, Ricketts TH. Farm and landscape factors interact to affect the supply of pollination services. Agriculture Ecosystems & Environment 2017; 250: 113-122.
Park MG, Blitzer EJ, Gibbs J, Losey JE, Danforth BN. Negative effects of pesticides on wild bee communities can be buffered by landscape context. Proc Biol Sci 2015; 282: 20150299.
Ponisio LC, M'Gonigle LK, Kremen C. On-farm habitat restoration counters biotic homogenization in intensively managed agriculture. Global Change Biology 2016; 22: 704-715.
Potts SG, Imperatriz-Fonseca V, Ngo HT, Aizen MA, Biesmeijer JC, Breeze TD, et al. Safeguarding pollinators and their values to human well-being. Nature 2016; 540: 220–229.
Pywell RF, Heard MS, Woodcock BA, Hinsley S, Ridding L, Nowakowski M, et al. Wildlife-friendly farming increases crop yield: evidence for ecological intensification. Proc Biol Sci 2015; 282.
Requier F, Odoux J-F, Tamic T, Moreau N, Henry M, Decourtye A, et al. Honey bee diet in intensive farmland habitats reveals an unexpectedly high flower richness and a major role of weeds. Ecological Applications 2014; 25: 881-890.

In Finland we are studing pesticide residue levels and their effects on honey bees and natural pollinators in Finnish agriculture (filed conditions) when the pesticide application is compliant. Compliant application means following the instructions given by the athority (Finnish Safety and Chemicals Agency Tukes).

The results of the study provide information about the pesticide exposure levels of pollinators in Finnish agriculture. Moreover, the results enhance understanding on the effects of the residues on wild pollinators in boreal farmland. These results will help to estimate whether the current pollinator protection measures in Finland are adequate.