by Pierre Mineau (Pierre Mineau Consulting) and Susan Kegley (Pesticide Research Institute)
Because the dose levels used in this study have already been the subject of criticism, we review known exposure levels so as to provide a comparison of concentrations in nectar and water sources encountered under actual field conditions with those used in the study.
The study was conducted as follows:
- A total of 18 colonies across three different apiaries were included in the study. All apiaries were managed identically in terms of treatment for mites and Nosema, with all hives in the study treated for Varroa with a formic-acid-based miticide and for Nosema with Fumagillin.
- Treated bees (12 hives) were fed with a half-gallon per week of a sucrose or high fructose corn syrup (HFCS) solution (sucrose concentration was not stated, but typical summer feed is 50% sucrose) with a concentration of 136 µg/L of imidacloprid (6 hives) or clothianidin (6 hives) from July 2 to Sept 17th, 13 weeks. The control group (6 hives) was fed the same sucrose solution without added pesticides.
- The feed was completely consumed by the colonies at the end of every week, which indicates the bees either consumed or stored all of the syrup. Assuming 50,000 bees per hive, the researchers estimated the dose of imidacloprid or clothianidin at 0.74 ng/bee.
- The bees were also allowed to forage freely. This means that they didn’t necessarily eat all of the treated sugar water right away—they may have made honey out of it and stored it for later use, during the winter (which is when the treated hives died).
- Hive strength was tracked over time to assess population changes.
Results of the study show that half of the treated hives (6 of 12) failed to survive through the winter; in contrast, only a single control hive (1 of 6) was lost. As interesting as the differential loss of colonies was the observation that accompanied these losses. In the author’s words:
“The honey bee clusters in the six surviving neonicotinoid- treated colonies were very small, and were either without queen bees, or had no brood.”
Also, the authors make it quite clear that the effect was NOT the result of the Varroa mite. This is undoubtedly one of the most noteworthy results of the study because it calls into question the hypothesis that the Varroa mite is the primary cause of the empty-hive syndrome that has been dubbed “CCD”.
“We found no significant difference in the degree of Varroa mite infection between the control and neonicotinoid-treated colonies. The average mite counts were 10-12 per 150 bees in the control and neonicotinoid- treated colonies, respectively, as assessed in mid-August 2012 (table 1). We later reduced the mite counts in all colonies to 1-2 mites per 150 bees after the applications of Miteaway Quick strips, a commonly used medicinal treatments prior to the arrival of winter in which it significantly reduced mite counts from 10-12 to 1-2 mites per 150 bees, respectively, in control, imidacloprid, and neonicotinoid-treated hives (paired t-test, p < 0.0001).”
Finally, the symptoms of colony death were very different between the control and treated hives. The original article provides photographs that show the dramatic difference between treated and control hives, with the treated hives nearly empty of bees and the control hive containing dead bees several inches thick on the bottom board.
“It is imperative to emphasize that while pathogen infections are common and serious diseases found in honey bees that often lead to colony death, the post-mortem examinations of the pathogen- caused dead colonies are vastly different to those suffered from CCD (Anderson and East, 2008; Lu et al., 2012). One of the defining symptomatic observations of CCD colonies is the emptiness of hives in which the amount of dead bees found inside the hives do not account for the total numbers of bees present prior to winter when they were alive (figure 3). On the contrary, when hives die in the winter due to pathogen infection, like the only control colony that died in the present study, tens of thousands of dead bees are typically found inside the hives (figure 4). The absence of dead bees in the neonicotinoid-treated colonies is remarkable and consistent with CCD symptoms.”
The response from Bayer Corporation, the major manufacturer of neonicotinoid insecticides, called into question the dosing regime used in the study:
“Bayer scientists have reviewed the study and, in their opinion, find it to be seriously flawed for major reasons. The Bayer scientists listed their concerns in the three following bullet points:
- “Feeding honey bees levels of neonicotinoids greater than 10 times what they would normally encounter is more than unrealistic—it is deceptive and represents a disservice to genuine scientific investigation related to honey bee health.
- Given the artificially high levels tested over 13 consecutive weeks, the colony failure rates observed are completely expected.
- Unfortunately, this latest study conducted by Dr. Lu repeats the fundamental flaws seen in his previous research and provides no meaningful information regarding honey bee risk assessment.”
But does Bayer’s claim that dose levels were 10 fold higher than expected stand up to scrutiny? Let’s take a closer look.
Lu and colleagues used solutions of neonicotinoids at concentrations of 136 µg/L. This concentration of pesticide in the solution is indeed higher than one would typically find in nectar in plants grown from treated seed. Based on industry field trials, the European Food Safety Authority (EFSA) estimates that the upper 90th percentile estimate of nectar concentration from neonicotinoid-treated seed standardised to 1 mg of active ingredient per seed (an average treatment rate) is about 42 ug/L. However, exposure as a result of seed treatments is not the only route.
In a recent article co-authored by scientists at UC Riverside and Bayer (Byrne et al. 2013), imidacloprid concentrations in nectar were measured over a range of 2.9 to 39.4 µg/L in citrus treated with a drip application of 560 g a.i./ha, currently the maximum rate permitted. Two other metabolites (both of which are of roughly equivalent toxicity to bees) were included in the analysis and found to be present, increasing the total concentration of all toxic residues by 1.5 to 1.7 fold. This brings the total toxic residue levels to approximately 4.6 – 63 µg/L using an average multiplier of 1.6. Samples of nectar collected from individual bees feeding exclusively on treated citrus flowers were as high as 37 µg/L. Based on the graphs presented in the paper, total residues in nectar were higher still after a fall application to citrus, with sampling occurring a full 230 days after application. When the nectar was evaporated by the bees to make honey, pesticide concentrations increased, with freshly capped honey having an approximately 4-fold higher concentration than the nectar removed from the bees’ crops when it was first brought into the hive or the flower nectar from treated trees.
So, while the concentration of the dosing solution used in the Lu study was somewhat higher than concentrations that would be encountered by bees either from common seed treatment or drip applications in citrus, it was not higher by a factor of 10. The concentration used by Lu and colleagues (136 µg/L of imidacloprid) is comparable to the maximum sample value seen by Bayer in their joint citrus study with the University of California (Byrne et al. 2013) in freshly capped honey (95.2 µg/L). Assuming that Lu used a 50% sucrose solution and that this was transformed into honey at 82% sugar concentration, the finished honey in Lu’s hives would contain a concentration slightly more than two-fold that of the higher value reported by Byrne in treated citrus.
More importantly, contaminated nectar is not the only source of exposure to neonicotinoids and other systemic insecticides. Average concentrations of imidacloprid, thiamethoxam and clothianidin in guttation droplets following seed-treatment of corn seedlings have been measured between 7,000 and 346,000 µg/L, based on work by Girolomi et al., clearly eclipsing the concentrations in nectar. Bees have been observed drinking guttation droplets, as well as from puddles and saturated soils, although how frequently they use these water sources is not well documented. Anecdotal observation by one of us (S.K.) suggests that bees will take water from moist soil in preference to an equidistant pond.
Guttation fluid from seedlings can contain very high concentrations of insecticide, much higher than those used in the Lu et al. experiment. Concentrations decrease over time as the plant grows, but remain a low-level source of exposure throughout the growing season. Photo credit: Bee Life
Bees appear to use moist soil as a preferred source of water. Puddles or moist soils contaminated with neonicotinoid insecticides can thus serve as another source of exposure. Photo credit: Susan Kegley.
Dose Per Bee Is Substantially Below the LD50
Because of the nature of the Harvard experiment, it may be more reasonable to compare exposure and effect in terms of the dose received by bees. Lu and colleagues estimate that they provided every hive bee with 0.74 ng/bee/day. This dosage is below the oral LD50 of 4.0 and 3.7 ng/bee oral toxicity values cited by European regulatory authorities for clothianidin and imidacloprid, respectively. (Lu and colleagues cite higher LD50 values, especially for imidacloprid, reflecting the wide range of values obtained from variable study conditions).
In the Byrne article, University of California and Bayer scientists arrive at the following:
“Assuming 6 h of flying per day, a forager would consume between 33 and 88 mg of sucrose. Using the sucrose concentrations estimated in the tunnel studies (25%) (Fig. 6), this corresponds to 133–350μL of nectar, a volume that would contain approximately 1.33 to 3.50 ng of imidacloprid.”
The authors of this article further state that “these amounts are well below published LD50 and NOEL values.” Whether or not one agrees that a predicted dose of 3.5 ng of imidacloprid is well below the 3.7 ng/bee LD50 value initially reported by Bayer and adopted by European regulators, it is clearly higher than the dose level of 0.74 ng/bee/day calculated by Lu and colleagues. Therefore the 0.74 ng/bee/day Lu et al. used in their experiment may be at the low end of what might be expected in citrus orchards treated with imidacloprid drenches.
It is true that flowering only lasts for about a month in citrus, so the exposure duration may be lower than the 13 weeks used by Lu and colleagues. However, if substantial amounts of nectar are stored in the hive as honey for future use, this difference in timing is moot. The actual daily consumption of insecticide by the bees over the course of the Lu study remains unknown. It is unfortunate that the authors did not quantify residues in finished honey. This would have allowed for a direct comparison with Byrne and others. Also, because the bees were allowed to forage freely, they had access to additional sources of nectar that may have been free of pesticides or may have contained unknown amounts of various insecticides and fungicides. A discussion on the environment of the apiaries and identification of the apiaries in which colony failures occurred would provide clarity regarding potential additional exposures.
When nectar is abundant, normal colony behavior is to transform nectar into honey, which is stored for later use. Thus, it is likely that the bees transformed at least some of the pesticide-contaminated sugar solutions into honey, rather than consuming it immediately. If this were the case, the fact that the control and treated hives did not show any significant differences in population during the summer months, but crashed during spring buildup when they tapped into their stored honey supplies would not be surprising.
We conclude that Bayer’s claims that this new research is “deceptive and represents a disservice to genuine scientific investigation related to honey bee health” is lacking in credibility. We believe that the work of Lu and colleagues needs to be examined and debated based on sound science and valid risk assessment assumptions.
About our guest blogger
Pierre Mineau is a retired government senior scientist and, currently founder and principal of Pierre Mineau Consulting. He is a well-published PhD scientist with over 35 years of ecotoxicology and ecological risk assessment expertise. He frequently advises non-governmental organisations, national governments and international associations on pesticide issues.