Here we show, for the first time, that the toxicity of a pesticide formulation to bees is caused exclusively by a co-formulant (alcohol ethoxylates), rather than the active ingredient. A 0.8 µL acute oral dose of the agricultural fungicide formulation Amistar® caused a range of damage to bees: both lethal, with 23% mortality, and sublethal, with 45% reduced sucrose consumption, 3.8% drop in body weight (whereas the negative control gained 4.8%), and a 302% increase in gut melanisation. For all metrics tested, the Amistar® and alcohol ethoxylates treatments were not statistically different, demonstrating conclusively that the toxicity of the formulation, Amistar®, to bumble bees is driven by the alcohol ethoxylates. These results demonstrate gaps in the regulatory system and highlight the need for a greater research focus on co-formulants.
The mortality in the Amistar® treatment, and treatments containing alcohol ethoxylates reached 32% at its highest, which is substantial given that bees are likely to have a high level of exposure to Amistar® and alcohol ethoxylates. The mechanism by which the alcohol ethoxylates cause mortality has not been explicitly isolated, but our results suggest two potential, possibly related, causes. We recorded a 302% increase in the melanised area of bee midguts in the alcohol ethoxylates treatment. A similar effect was observed in Melipona scutellaris exposed to the pure fungicide active ingredient pyraclostrobin alongside a similar reduction in survival37. We suggest that the alcohol ethoxylates are disrupting the structure of the midgut, which the bee immune system is reacting to with melanisation44 (see Fig. 5). In parallel with this gut damage, alcohol ethoxylate treatment drove a 54% reduction in sugar consumption, which persisted throughout the experiment. Supplementary Fig. S3 shows a plot comparing sugar consumption against gut melanisation, with increasing gut melanisation correlated to reduced sugar consumption in the Amistar®, co-formulant mixture and alcohol ethoxylates treatments. Consequently, we propose that mortality was driven by energy depletion due to reduced consumption, which in turn may have been driven by damage to the gut.
(Left) Bumble bee midgut in the negative control treatment. (Right) Bumble bee midgut in the co-formulant mixture treatment, which contains alcohol ethoxylates. The dark brown patches are areas of melanisation, indicative of damage to the gut. Both bees survived the full 120 h.
Likely as a consequence of the reduced consumption of sucrose, bumble bees in the alcohol ethoxylates treatment lost 8.4% of their original weight, in stark contrast to the negative control where bees gained 4.8% over the five-day period. This indicates the alcohol ethoxylate treated bees were expending more energy than they were consuming, and thus exhibiting a negative energy balance. This weight loss, while considerable as a percentage of the bee’s total body mass, is also similar in scale to the weight of the sucrose bees consume in one sitting (EA Straw pers. obs.), for which rigorous data do not exist. As such it is possible that a portion of the weight loss is attributable to the reduced sucrose consumption of the bees, meaning they would have less sucrose in their guts at the time of weighing. Sucrose consumption does not, however, explain the failure of alcohol ethoxylate treated bees to gain weight, which was observed in the control treatment. The weight loss, and lack of weight gain, are concerning because they are likely to indicate a reduction in fat reserves, although this has not been experimentally confirmed. Bee fat reserves are important physiologically, in particular in responding to immune threats45,46. Fat reserves allow bees the energetic resources to buffer against challenges, and thus their depletion could expose bees to greater risk from future threats47.
The reduced appetite and negative energy balance in alcohol ethoxylates treated bees could have broader effects in the natural environment. Bees pollinate flowers as they forage for nectar and pollen, so a reduction in their appetite could subsequently have effects on ecosystem services. In our experiment, bumble bee appetite was reduced immediately after ingesting a single dose of alcohol ethoxylates or Amistar®. This effect persisted for five days after exposure, indicating a persistent change in consumption behaviour. While nectar-foraging in bumble bees is driven by the needs of the colony48, a reduction in appetite would reduce overall colony nectar consumption, and thus the number of foraging trips made for nectar. Fewer visits to flowers for nectar may lead to reduced pollination, which would be detrimental to crop yields and farm profits. Further studies of how the impacts we have found map onto foraging and pollination are clearly needed. Importantly, the reduction in appetite recorded in our experiment is a sublethal effect, which standard lower tier testing would not detect. When Amistar® is tested on bumble bees for the 2025 renewal of azoxystrobin, this sublethal effect will be missed by regulatory testing, despite the impact it may have on the pollination services such testing is designed to protect. We suggest that a simple modification to the regulatory protocol OECD 247 would be to weigh the sucrose syringes at the start and end of the trials to calculate sucrose consumption, which would allow measurement of this sublethal effect with minimal additional workload.
Our results show a slightly, but not significantly, higher level of mortality in the alcohol ethoxylates treatment (30%) than the Amistar® treatment (23%). If this is a real biological difference, one explanation might be that the concentration of alcohol ethoxylates in the Amistar® formulation was lower than that used in the alcohol ethoxylates treatment solution. This is possible because the Amistar® material safety data sheet lists concentrations as a range (10–20% for alcohol ethoxylates), and here we used the upper end of the range. The co-formulant mixture treatment in all metrics was statistically indistinguishable from the alcohol ethoxylates treatment, showing that the toxicity of alcohol ethoxylates is not a result of synergism with other co-formulants.
We believe that the implications of our results are not limited to a laboratory setting and a single species, as other published and unpublished research supports our findings. Semi-field flight cage experiments, where Amistar® was applied to a crop, found effects on full bumble bee colonies (Bombus terrestris). Amistar® caused a reduction in average bee weight and a reduction in foraging activity, as our results predict49. This demonstrates that the effects observed in our laboratory testing scale up to effects at a field realistic level. Additionally, in honeybees (Apis mellifera) Amistar® has been found to cause mortality in laboratory experiments at a range of doses50,51, demonstrating the mortality effect found in our experiment is not species specific. However, no mortality was seen in trials on the red mason bee Osmia bicornis (Hellström and Paxton, unpublished data). Additionally, a similar compound, C11 and lower alcohol ethoxylates, has been found in small scale laboratory testing to cause 100% mortality after contact exposure in honeybees31.
To measure the exposure of bees to PPP’s, the EU mandates trials that measure chemical residues in pollen and nectar after crops have been sprayed with either active ingredients or formulations34. However, these residue analysis studies only measure active ingredient concentrations, not the co-formulants. As such, we have no systematic data on the exposure of bees to co-formulants7,8,9. This dearth of data means that the exposure of bees to co-formulants is very poorly characterised. To estimate exposure to alcohol ethoxylates, residue data for Amistar®’s active ingredient azoxystrobin could be used as a proxy18,52. However, the chemical properties of alcohol ethoxylates, specifically their surfactant action, make it unlikely that they have an equivalent environmental fate to azoxystrobin, so this would not be appropriate.
While we have very little data to quantify bee exposure to alcohol ethoxylates, we know Amistar® can be applied to crops, such as strawberries, during flowering while bees are foraging on them. The Environmental Information Sheet for Amistar® states “[For bees] no risk management is necessary. Amistar® is of low risk to honey bees”53,54,55. In addition, we would note that exposure of bees to alcohol ethoxylates, and related substances, is not exclusively from Amistar®. For example, a cursory search of the Syngenta website56 immediately identified alcohol ethoxylates in five other Syngenta products. Worryingly, the chemical group alcohol ethoxylates sit in, alkoxylated alcohols, are also widely used in adjuvants, which are products which can be added to tank mixtures to modify the action of the agrichemical6. 89 adjuvant products licenced in the UK contain alkoxylated alcohols as the primary ingredient15. To our knowledge, these adjuvants have never been toxicity tested on bees and have no bee exposure mitigation measures in place whatsoever.
To complement measures to promote academic research, moving regulatory research beyond its mortality and active ingredient-centric approach to toxicity testing would better reflect the risks pesticides, as used in the field, pose. For regulatory systems to accurately characterise risk they need to estimate the scale of sublethal effects, regardless of initial mortality results33. The results presented here demonstrate that even substances assessed by regulators as ‘bee safe’ can pose a serious hazard to bee health. To reflect potential sublethal differences caused by co-formulation composition, all formulations could undergo a much more rigorous set of lower tier testing or be automatically entered for higher tier testing.
In the face of declining bee populations we advocate that a precautionary approach minimising the exposure of bees to potential stressors, where possible, would be prudent. The current legislation allowing application of PPPs directly onto bees and flowering plants does not align with the emerging evidence that co-formulants, adjuvants, herbicides and fungicides can be hazardous to bees8,57. The wealth of untested and undisclosed co-formulants used abundantly in agriculture is a serious and pressing concern for the health of pollinators worldwide.
Source: Ecology - nature.com
