Additional Comments on the Issue of Neonicotinoid Seed Treatments and Bee Mortality

Since posting my early September blog ( about neonicotinoid seed treatments and bee mortality, the Pest Management Regulatory Agency (PMRA) of Health Canada has issued a Notice of Intent on “action to protect bees from exposure to neonicotinoid pesticides” ( The following comments pertain to that notice and a related PMRA report on bee deaths in Ontario and Quebec in 2012, some additional data they’ve collected for 2013, other published research papers/reports, and further discussions with Canadian bee researchers and professionals. 

1.         The PMRA Notice of Intent states that “the majority of pollinator mortalities were a result of exposure to neonicotinoid insecticides, likely through exposure to contaminated dust generated during the planting of treated corn seed.” And it goes on to say, “we have concluded that current agricultural practices related to the use of neonicotinoid treated corn and soybean seed are not sustainable.” The Notice of Intent seems to be based largely on information contained in a PMRA report entitled, Evaluation of Canadian Bee Mortalities Coinciding with Corn Planting in Spring 2012 (not on web).The 2012 report involves an investigation of reports of bee deaths from 40 beekeepers involving a total of 240 hive locations in Ontario and one report involving eight hive locations in Quebec. 

2.         PMRA tested, for neonic residues, 125 samples of dead and 2 samples of “dozy” bees from 25 affected beekeepers, plus 20 samples of living bees from apparently healthy hives. The sampling occurred sometime after corn planting, but before corn pollination. The PMRA’s finding that 70% of the dead bee samples in Ontario had detectible residues of chlothianidin, while residues were detected in only one “unaffected” bee sample, has received major attention. (Note that clothianidin is the active ingredient in the seed treatment commonly known as ‘Poncho,’ and is also a breakdown product of thiamethoxam, or ‘Cruiser,’ the other commonly used corn seed treatment.) 

3,         An appendix table in the “2012 deaths” report, listing neonic residue levels for the 127 dead- and dozy-bee samples, is of major significance. PMRA states that the 48-hr “NOELs” (No Observable adverse Effect Levels) for clothianidin are 0.0085 (oral exposure) and 0.067 ppm (dermal exposure). NOEL is the minimum exposure level considered to have any negative effect on bees. If it’s contact with neonic-laden dust in air at planting time, dermal would seem like the most likely means of exposure, with oral exposure being more appropriate for neonics in floral nectar and pollen. Importantly, data in the appendix table show that not one of the 127 dead/dozy bee samples had neonic residue concentrations as high as the NOEL dermal level. The highest sample reading was 0.024 ppm. Only 17 of the127 were even at or above 0.0085 ppm. 

In summary, 70% of the dead samples had detectable neonics, but 100% of samples were below NOEL values. 

PMRA’s report says that residue concentrations in dead bees may decline with time, because of bacterial decomposition. Sunshine exposure also breaks neonics down. However, experts in the field say that the rate of decline in dead bees is largely unknown. Hence, it’s not clear whether the amount of neonics in any of the dead bee samples was ever high enough to cause adverse effects to these bees while they were still alive – at least based on available science. 

4.         By contrast, what is well established scientifically is that the rate of breakdown of neonics in living bees is very rapid. I’m told that the half-life for clothianidin in bees is 11 hr, or less. After 3 days, only about 1% of the original exposure remains. It’s no surprise that PMRA did not detect neonics in most living bee samples presumably collected several days after incident reports were submitted. This perspective seems to be missing from both the PMRA 2102 report and Notice of Intent. 

5.         A summary (not published) of 2013 PMRA analyses of neonic levels in dead bees does not give the breakout for individual samples, only the range, which is from 0.001 to 0.071 ppm, with 75% of all dead bee samples having detectable clothianidin. The 0.071 is close to the NOEL dermal limit (0.067 ppm). Hence, one sample out of more than 225 over two years was as high as the minimal level known to harm bees. 

6.         The PMRA’s statistic that 70% of dead bees (and 75% in 2013) had detectable neonic residues has also much to do with the precision of the detection technology. The “limit of quantification” in the PMRA analyses is 0.001 ppm. If it had been 0.01 ppm, the 70% in 2012 would have been 12%. And if the limit of quantification had been 0.000l ppm, the 70% figure would almost certainly been even higher. With sufficiently sensitive technology, given the wide use of neonics in modern society for uses well beyond corn seed treatment, almost 100% of the samples might have been expected to have neonic residues. The critical issue is not whether neonics are detected but whether they are present at levels which are injurious to bees. The 2012 document implies “no.” Preliminary 2013 data say “maybe in one case.” 

7.         PMRA’s identification of soybean seed treatments as a probable cause of bee deaths is most puzzling, especially when the agency goes so far as to say, “agricultural practices related to the use of neonicotinoid treated … soybean seed are not sustainable.” But PMRA has not presented any direct evidence linking bee deaths to soybean seed treatment. All of the discussion/analysis in its “Evaluation of Canadian Bee Mortalities Coinciding with Corn Planting in Spring 2012″ pertains to corn. Why did they do this? The rationale might have been: ‘if corn seed treatment is bad, then soybean treatment must be bad too’ (scarcely a word said about canola or other crops). 

Or maybe they assumed that soybeans are planted with corn planters with the same practices/problems in dust emissions. While it is true that some soybeans are planted with corn planters, the vast majority of the acreage is planted with equipment more akin to that used for canola. 

8.         Then there’s another enigma: The PMRA document about 2012 bee deaths contains a map showing the location of the 72 bee yards from which samples were analyzed. They are not uniformly distributed across the corn/soybean growing area of Ontario. In fact, about one-third of those yards are in a small pocket near Hanover, (Grey County) Ontario. While corn and soybeans are definitely grown there (canola too), this is not an intensive area of corn and soy production as a map in the PMRA report shows. The other site where a large portion of the samples were collected is Middlesex and East Lambton counties, an area which does grow a lot of corn and soybeans. 

9.         In short, while the PMRA document does make a modest case for bee deaths being caused by corn (though not soybean) neonic seed treatments, the supporting data are far from overwhelming. 

10.       Of interest, also, is an analysis which Cutler et al. (2012, ) did of PMRA’s “bee incident reports” for pesticide injury during the years 2007 through 2012. There were a total of 110 incident reports filed by Canadian beekeepers, 104 of them in 2012 alone. PMRA divided these incident reports into three categories, minor, moderate and major. Major incidents were defined as having at least 3000 dead bees from each of five or more colonies, or 30% of the bees in any one colony, dead or exhibiting abnormal behavioural effects. Of the 110 incidents, 78 involved neonics or a combination of neonic plus another pesticide. However, only four of the 20 major incidents involved neonics. In fact, five of the major incidents involved exposure to formic acid, an organic miticide that beekeepers apply to colonies to control varroa mites on bees. The formic acid incidents are especially notable for reporting queen bee deaths. Here’s another report linking bee colony mortality to miticide treatments, .

 11.       There is another part to this puzzle. Some vocal beekeepers claim that the effect of neonic seed treatment for corn has been a drastic drop in bee numbers in Ontario. However, when presented with data from Statistics Canada showing that colony numbers have actually been increasing at a steady rate in recent years, they say that the critical effect is on over-winter mortality, citing an average of about 35% mortality during the 2012/2013 winter. (Average Canadian mortality has ranged from 15% to 35% since 2006/07 with the highest percentage actually being in 2007/08.) 

But how can an acute effect at corn seeding time in late April-May affect the mortality of bees the next winter? This is especially puzzling given the rapid rate at which neonic chemical are broken down by living bees, as well as being photo-decomposed when exposed to daylight. And with worker bees only living an average of about 40 days, over-winter deaths would involve bees born months after the seeding-time exposure. One explanation might be neonic effects on queen bees which live much longer; however bee experts tell me this is highly unlikely given the filtering which occurs with queen bee feeding. (Queen bees eat “Royal jelly,” a special type of food manufactured by other bees.) Queen bees will also break neonic chemicals down quickly in the unlikely event of significant exposure. 

12.       Re-exposure to neonics later in the growing season is another possibility. “Let’s-ban-neonics” advocates have based their arguments on claims that the agricultural environment is full of neonics and that pollen from plants grown in neonic-containing soils is another cause of deaths. But the supporting evidence seems very weak. In published reports, where neonics have been detected in soil or plant pollen, the concentrations have been far too small to be biologically significant to bees. Bees would have to ingest daily amounts in excess of their body weights to approach the NOEL values referenced above. There is an excellent published report from southern Germany detailing how improper corn seed treatment led to spring-time bee deaths in 2008 ( (The amount of in-seed-bag dust per 100,000 seeds was 10 to 100 times higher than acceptable standards.) But even here the amount of chemical in subsequent corn pollen was too low to be significant. 

PMRA checked for neonic residues in Ontario farm soils and water puddles in 2013, with the residue levels being mostly “not detected” or at ultra-low levels. There is a report of neonics in surface water samples in Quebec but details are sketchy. 

In summary, statements that corn neonic seed treatments are responsible for over-wintering bee deaths make no sense, scientifically. 

13.       One big dilemma in policy decisions linked to neonic seed treatments is how to design  Best Management Practices (BMPs) based on PMRA data and policy dictates. The normal response would be to reduce exposure to a pesticide so that residue levels fall well below NOEL values. But we appear to be already there according to 2012 and 2013 PMRA measurements. So what’s the new target – 10% of NOELs, 1% of NOEL, or even lower – knowing that absolute zero is impossible? And if we switch completely to other pesticide products for seed/seedling insect control, what assurance is there that the resultant environmental problem will not be worse. (The reason for going to neonic products to begin with largely involved safety to humans and environment with other compounds.) 

14.       Some work done by Dr. Krupke and colleagues in Indiana is of interest – research results which are often cited as “proof” that neonic seed treatments kill bees. Krupke et al. (2002, reported ultra-high levels of neonics in talc dust collected on the exhaust manifold of their pneumatic corn planter; they and others have used this to emphasize the seriousness of neonic exhaust from pneumatic corn planters. But as others have pointed out, this says little about emission rates. (It’s like scraping soot off the inside of a tailpipe to estimate auto exhaust emissions.) Perhaps more significantly, they placed bee colonies on all four sides of a small field being planted with this corn planter. While the research paper is silent on the fate of the bees, Dr. Krupke has told others that there were no apparent toxicological problems with these bees. (The paper has other weaknesses, but they’re beyond the scope of this blog.) 

15.       And that leads to a final question: What’s causing the bee deaths? 

There is no doubt in my opinion that there are incidents where using a pneumatic vacuum-style corn planter, with talc added to neonic-treated seed, and with wind blowing strongly towards a concentration of bees (hives, pollinating flowers) at seeding time, that neonics will kill bees. 

But I’m hearing from bee researchers that a larger cause of bee mortality is bee diseases, and most notably virus diseases spread by varroa mites. 

Some beekeepers have claimed, “This can’t be my situation because my varroa levels are low.” But the scientists respond that it takes very few varroa to introduce critical viruses into a bee hive. Once there, they spread quickly – and even to new bees introduced many months later into supposedly empty, though infected, hives. Both Dr. Rob Currie at the University of Manitoba and Dr. Ernesto Guzman, University of Guelph, have research (as yet unpublished) which implicates varroa-transported viruses as being an important cause of abnormal bee deaths. Few viral measurements have ever been taken for commercial bee hives in Ontario, and most viruses are difficult/impossible to detect visually. Bees dying from viruses display symptoms similar to those dying from pesticide exposure. (Canada has only now established its first bee virus lab service. It’s at Beaver Lodge in Northern Alberta, two days away from Ontario and Quebec by courier yet the bees must arrive alive, and the per sample cost is $325.) 

And there may be another factor too. At least one vocal Ontario beekeeper has reported that his bees have done better once he moved them “further north” where the proportion of land planted to corn and soybeans is lower. He also said the “problem” has become more notable in the last few years. He attributes this to neonics, but there could be another cause – bee starvation/malnutrition. There is little (soybeans) or no (corn) nectar with these crops. Corn provides pollen for protein though I’m told bees only collect it if other more attractive sources are not available. With recent higher grain prices, many former forage (i.e., alfalfa/clover) fields have recently been planted to grain crops. Bees positioned near these fields may have had little to eat, especially if many dozens of hives are located at one location, with each hive holding up to 80,000 worker bees. It’s a case of not enough food and too much competition.

 16.       I generally support new OMAF guidelines for corn seed treatments (, and welcome Bayer’s plans for a safer alternative to talc powder. Planter manufacturers could help. German research showed emissions into surrounding air dropped by 90% plus when exhaust was directed to the ground. 

Suggestions that only 20-30% of corn seed needs insecticide treatment make me uneasy. Past field experience is not that useful if you’ve been using treated seed. Agricorp (Ontario’s crop insurance agency) requires a backup insect control plan if using untreated seed, to ensure crop insurance coverage in Ontario. 

I don’t believe corn growers should seed neonic-treated corn using vacuum planters with wind blowing toward areas of bee concentration. Options include different planters, untreated seed, better seed lubricants, calm winds, and/or moving the hives.