Critique: Multiple Routes of Pesticide Exposure for Honey Bees Living Near Agricultural Fields.
Table of Contents
1.0 INTRODUCTION TO ARTICLE.........................................................................................................................3
3.0 CORE CRITICAL REVIEW...................................................................................................................................7
MATERIALS AND METHODS.................................................................................................................................................9
1.0 Introduction to Article
Multiple Routes of Pesticide Exposure for Honey Bees Living Near Agricultural Fields is a
scientific paper that attempts to explain some of these routes of exposure and tries to find a direct
correlation between honeybee mortality and the use of various forms of neonicotinoid pesticides.
The study was conducted by the Department of Entomology at Purdue University in West
Lafayette, Indiana at the Purdue Agricultural Center. The authors are C.H. Krupke, G.J. Hunt,
B.D. Eitzer, G. Andino, and K. Given. The editor was Guy Smagghe from Ghent University in
Belgium. The article was published on January 3, 2012 by PLoSONE at www.PLoSONE.org,
which is an open source scientific publishing website that only publishes scientific articles that
have not been published elsewhere and represent the results of primary scientific research only.
Honey Bees are one of the most important pollinators in the world and are crucial for the world
food supply. Worldwide there have been numerous reports of unexplainable bee deaths whereby
bees did not return to the hive after foraging and either starved to death or couldn’t find their way
back and died of exposure. The dead bees, pollen, and wax from these hives routinely contain
detectable levels of insecticides, fungicides, miticides, and herbicides.
Neonicotinoids have been heavily adopted in both agricultural and urban environments for their
low vertebrate toxicity and the ability to be translocated by plants (Jeschke P., 2008). The single
largest application of neonicotinoids is as a seed coating on maize seed. In 2010, maize planting
reached unprecedented levels in the US and is expected to increase. Almost all maize seed
planted in North America has neonicotinoid seed coatings. Current seed coating application rates
range from 0.25 to 1.25 mg/kernel. The LD 50 for Honey Bees ranges from 22-44 ng/bee for
clothiandin. One seed contains enough neonicotinoids to kill 80,000 bees (Krupke, 2012).
The study was done in response to reports of major bee kills in Indiana in 2010 that coincided
with maize planting. Neonicotnoids are known to be highly toxic to honeybees, however,
because they are not the target organism, little is known about the multiple routes of exposure.
The study attempts to understand the various routes of exposure of neonicotinoids and asks
several questions “why are bee deaths greatest at planting time?; what are the routes of
exposure?; and how are farming practices contributing to this problem?”. This paper attempts to
answer these questions and concludes that the seed coating combined with talc plays a significant
role and is the major cause of bee deaths.
It is widely accepted that neonicotinoids are toxic to honeybees (Horan, 2010), but it can be
argued that when used as recommended, it is safe. However, previous studies did not take into
consideration the addition of talc to the seed bucket, which is added to prevent seeds from
sticking together and results in more efficient and evenly spaced plantings. This study was the
first to analyze fugitive dust emissions produced from the residue of talc and seed coating that
gets dispersed when planting.
The study began in the spring of 2010 during the planting season and incorporates two years
worth of data. It was initiated in response to bee kills in Indiana during the spring of 2010 that
coincided with maize planting. Analysis of dead bees, live bees, and pollen revealed that both
clothiandin and thiamethoxam were found in dead bees and pollen from a single hive, however,
these compounds were not found in live bees. After these preliminary results, the study was
expanded to determine potential routes of exposure of these pesticides to Honey Bees throughout
the season, both during planting and later during plant development.
Neonicotinoid pesticides act as nicotine acetylcholine receptor agonists in invertebrates. These
targets play an important role in bee learning and memory processes vital to successful foraging
behaviour (Gauthier, 2010). Two major compounds used in neonicotinoids are clothiandin and
thiamethoxam. Thiamethoxam gets metabolized to clothiandin by the insect. Both compounds
are known to be highly toxic to honeybees, causing memory loss, confusion, and death.
The experiment was conducted in 2010-2011 at Purdue University’s Agricultural Center, which
is comprised of agricultural land that is owned and maintained by Purdue University.
Before planting, soil samples were taken from the top 2-3 cm of soil at four equidistant locations
in each field were homogenized, and analyzed. Eight hives were placed on the perimeter of an
unplanted field that was surrounded by areas of long-term maize and soybean planting. Pollen
traps were placed at the entrance to each hive and pollen samples were collected daily before and
after the field was planted.
During planting, half the field was planted with commercial maize seed treated with
1.25mg/kernel of clothiandin and talc was added to the see box at the recommended rate. The
other half of the field was planted with untreated seed and talc was added at the recommended
rate. Waste talc was collected after each planting from the planter vacuum system and analyzed.
The planter box and vacuum system were exhausted and thoroughly cleaned with compressed air
before each planting. This cleaning was done at a location far from the experimental field.
Samples were collected and analyzed for pesticides from potential exposure routes from these
fields. Samples included soils, pollen collected directly from the plant, pollen collected from
bees, dead and healthy bees, dandelion flowers, and waste talc left in the seed box after planting.
Fresh pollen from the anthers of 100 maize plants per field was collected by removing the anther
during pollen shedding and placing them into a bag to separate the pollen from the anther.
Sampling was conducted in 2011 after another report of dead and dying bees at a local apiary,
which coincided with planting. Bees were collected from several hive entrances and analyzed.
Also, apparently healthy returning foraging bees from the same hives were collected and
analyzed. Nectar and pollen were collected and analyzed from the frames of two hives, one with
dead bees at its entrance and one from a nearby hive without any visible dead bees. Surface soil
and dandelion flowers were collected from maize fields within 2 km of this apiary. The planting
of this field coincided with the collection of bees and samples. Additional waste talc samples
were collected from commercially available seed from various manufacturers. The same fields
were replanted using the same methods and materials as 2010.
Samples were prepared using the QuECHeRs protocol and analyzed using liquid
chromatography/mass spectrometry. The combination of these techniques allows for various
pesticides to be unambiguously identified at the parts per billion concentration level.
Pollen identification was required to identify maize pollen from other sources of pollen.
Reference slides were used to compare and identify maize pollen while online images were used
to identify the other pollen types. Four types of pollen were identified, maize, common dandelion,
plantain, and goldenrod. To estimate the proportion of maize pollen, a subsample was suspended
in water, dried and individual maize pollen grains were counted. The volume of 5 individual
grains from each subsample were averaged and used as the volume.
The results in 2010 indicate that all soil samples, tested had neonicotinoid residues, with
clothiandin found in every field (Figure 1). Results of waste talc samples from planting activities
revealed extremely high levels of clothiandin in talc exposed to treated seed (Figure 2). Analysis
of talc from untreated seeds revealed very low levels of the same pesticides and was attributed
contamination caused by inadequate cleaning of the equipment between plantings. Anthers
sampling revealed the presence of the same compounds but in far lower concentrations. Pollen
collected from pollen traps found thiamethoxam in 3 of 20 samples and clothiandin was found in
10 of 20 samples. By volume, 50% of pollen collected by Honey Bees was maize pollen.
Figure 1: 2010 Pesticide Concentrations from Surrounding Fields Soil Samples (ppb)
Figure 2: 2010 Pesticides Concentrations found in Waste Talc after Planting with Treated and
Untreated Seed (ppm)
The results in 2011 had the same trends, clothiandin was found in all dead and dying bees while
healthy bees from the same areas did not have detectable levels of clothiandin. These results
suggest that Honey Bees living near agricultural fields are exposed to neonicotinoids and other
pesticides via multiple routes of exposure throughout spring and summer, with the greatest
potential for exposure occurring during and after planting by the exhaustion of fugitive dust into
the environment. This fugitive dust is produced by the abrasion of talc with treated seeds,
causing the removal of the coating. This dust is collected by a vacuum system and released into
the environment between plantings.
A critical finding was the detection of stored pollen in cells and in pollen traps because
clothiandin is more toxic when administered orally. The methodology of the study does not
allow for the determination of oral toxicity or contact toxicity and may useful for further studies.
The findings of this study are of great significance to Canada. Canada employs the same farming
practices and uses the same pretreated seeds. The last few years have resulted in unusually high
numbers of reported bee mortalities, similar to the situation in the US. In Canada, virtually all
corn seeds and half of soybean seeds are pre-treated with neonicotinoids. Health Canada’s Pest
Management Regulatory Agency (PMRA) has concluded that current agricultural practices
related to the use of neonicotinoid treated corn and soybean seed are not sustainable and the
majority of pollinator mortalities were a result of exposure to contaminated dust generated
during the planting of treated corn seed. The PMRA issued additional protective measures for
corn and soybean production for the 2014 planting season, which included:
Requiring the use of safer dust-reducing seed flow lubricants;
Requiring adherence to safer seed planting practices;
Requiring new pesticide and seed package labels with enhanced warnings; and
Requiring updated value information be provided to support the continued need for
neonicotinoid treatment on up to 100% of the corn seed and 50% of the soybean seed.
Health Canada is working with industry partners to find more sustainable options while also
continuing the use of these seed treatments. Important research projects to improve integrated
pest management techniques are being discussed and potential solutions are being explored.
Health Canada's PMRA is currently expediting a re-evaluation of all uses of neonicotinoid
insecticides in cooperation with the United States Environmental Protection Agency (US EPA)
as part of the work being done with international partners. Health Canada's PMRA applies a
science-based approach to regulate pesticides, so any new information such as the findings of
this study are critical to Health Canada’s re-evalutation of neonicotinoids (Health Canada, 2013).
3.0 Core Critical Review
Overall, the paper is very detailed, lacked chronological order, and was difficult to follow. The
hypothesis and the research questions were not clearly defined and required re-reading to
formulate. The inclusion of the sources of funding and potential conflicts of interest provided
evidence that the researcher is unbiased and is not in a conflict of interest situation. Many
researchers receive funding from multi-national corporations, which may compromise academic
freedom and integrity.
The introduction was concise and provided important background information about
neonicotinoids, bees, and maize seeds. The results section followed the introduction, which made
it very difficult to understand without knowing how the experiment was performed. It would
have been easier to read if the methods section followed the introduction.
The results section lacked structure and flow and was very disjointed. Several references were
made to tables 1-6, however, these tables form part of the discussion section. As a result, the
reader has to keep flipping back and forth to understand. All tables should be listed together and
contained within the results section so that they can be easily compared with each other and the
The results section first describes the 2010 results and then the 2011 results, however, both are
from different experiments under different conditions, in different locations using different
methodologies. The 2011 results were collected from another location after a report of more dead
and dying bees at a local apiary in Indiana and is completely unrelated to this experiment and
should not form part of the results.
In the 2011 results description, wax combs were removed from two hives in the study apiaries
and both pollen and nectar were sampled for clothiandin and thiamethoxam. The pollen
contained both neonicotinoids but the nectar did not contain any and no explanation was given. It
is quite possible that the nectar was produced before planting and was not exposed to
neonicotinoids because the bees were consuming pollen and saving the nectar for the brood. This
is quite likely an unexpected result and should be explained or it may be an indication of errors
in the results or nectar sample collection that should have been addressed. It is highly likely that
if pollen contains neonicotinoids, the nectar should as well. However, if the results are accurate,
then further studies may be required. More samples should have been taken to support of refute
The results section also explained how dandelion flowers could be used as controls. There is a
negative correlation between distance from the planted fields and concentration of neocotinoids.
Dandelions growing far away from agricultural fields had non detect levels of neocotinoids while
concentration levels increased as distance from agricultural fields decreased (Krupke, 2012).
The tables are easy to read and summarize the results very well. One criticism is that table 2 used
parts per millions the units while the other five tables used parts per billion, which is inconsistent
and leads to confusion. All units should be the same to allow for quick and easy comparison.
The results are very supportive of the previously defined research questions and provide
evidence that using talc with seeds that have been pre-treated with neonicotinoids at the
recommended rates with the recommended equipment and procedures is killing honeybees.
The discussion section was very well written and detailed. It explained the results and connected
them to the methods used. There are numerous references and comparisons to previous studies.
Similar findings from other papers are compared as well as anomalies and unexpected results.
The paper lacked a well-written and definitive hypothesis. After reading the paper, a hypothesis
can be easily formulated by the reader. The research was done to determine the exposure
pathways experienced by honeybees near agricultural fields and the effects that farming practices
have on these multiple routes of exposure.
This study provides important information that supplements and supports other research dealing
with the decline of honeybees and Colony Collapse Disorder (CCD). More importantly, this was
the first to link the use of talc in treated seeds, current farming practices, neonicotinoids, and real
agricultural farming conditions, with CCD. It has been widely accepted that even low doses of
clothiandin are highly toxic but few have studied the effects on entire colonies of honeybees.
Until this study, all variables were studied independently under controlled conditions, which
greatly reduced the concentrations of pesticides to within sub-lethal limits resulting in a poor
understanding of CCD.
There is overwhelming evidence in primary literature that neonicotinoids are toxic to honeybees
at very low doses (Williamson S.M., 2013). Research by Chensheng (Alex) Lu and colleagues,
from the Department of Environmental Health at the Harvard School of Public Health released
the results of a study that was published in the Bulletin of Insectology in May, 2013 (Lu C.,
2012). This study provides additional data on the effects of neonicotinoid insecticides on bees,
and supports earlier studies demonstrating that exposure to neonicotinoids can compromise
honey bee survival over the long term. (Helman, 2013).
A comparison was made to a previous study that used canola instead of maize. The results found
low levels of clothiandin in pollen and nectar and concluded that there were no significant effects
on honeybee populations from clothiandin (Cutler GC, 2007). Based on the findings of the
current study, there was ten times more clothiandin found in maize pollen than canola pollen and
50% of all pollen collected was maize pollen. The researchers concluded that maize pollen was
the preferred food source over canola, which would result in more exposure to neonicotinoids
and higher mortality rates. This is a significant finding and is consistent with other papers that
have found that bees prefer maize pollen to canola pollen. Bayer Crop Science uses canola in
many of their studies, and have been criticized for misleading and inaccurate scientific research
(Schwartz, 2010). These results are more representative of actual clothiandin exposure levels to
The effects of neonicotinoid dust from maize seed-dressing on honey bees was recently studied
in Italy (Sgolastra, 2012). This study confirms the result of this paper, however, it was performed
in a controlled lab setting and did not reflect in-situ field conditions.
Although much of the research quantifies toxicity, the paper emphasizes the fact that there has to
be a pathway for exposure to occur. The intersection between toxicity and exposure is the key to
determining how much risk there is to honeybees from neonicotinoids.
The conclusion was included as part of the discussion section. The results of this paper suggest
that used talc exhausted during and after planting stands out as potential routes of exposure that
should be prioritized for further quantification and remediation. Producers are largely unaware of
this problem and its toxicity to honeybee. Priority should be given to the creation of guidelines
and procedures that prohibit waste talc from being exhausted into the atmosphere.
The conclusions at the end of the discussion relate back to the original research questions and
emphasize the importance of these results for honeybees and other pollinators that forage in
Materials and Methods
The materials and methods used in the study were described in great detail and were consistent
with that used in other similar studies. Equipment and protocols were clearly stated.
The inclusion of an Ethics statement attests to the researchers integrity and purpose; to provide
the most realistic simulation of real farming practices in order to gather the most accurate data
available. Many other studies do not include an Ethics statement. No permits were required and
no endangered of protected species were involved. All products used for maize planting were
registered for legal use and applied according to label guidelines. This statement emphasizes the
fact that the product was used and applied as the manufacturer recommended, which can be used
as a form of quality control as application rates and techniques may vary depending on the
operator of the equipment. It is a reflection of the researchers commitment to further the
understanding of honeybee CCD.
All samples were analyzed using a modified version of the QuECHeRS protocol. This method is
well established and widely accepted in scientific literature. Developed in 2003, it is a fast, easy,
and inexpensive method for the determination of pesticide residues in fruits and vegetables.
Pesticides are extracted from the sample by agitation with water, acetonitrile, and the salts of
magnesium sulphate and sodium acetate. The sample is centrifuged and the supernatant, which
contains the pesticide residue, is further cleaned by the addition of a primary amine, magnesium
sulphate, and C-18 silica. An aliquot of solvent is then analyzed using liquid
chromatography/mass spectrometry. The method provides both qualitative and quantitative
information in that it identifies various pesticides at the parts per billion level. (Anastassides M,
Although pollen identification was probably not an activity that made a significant contribution
to the results, the method used was quite subjective and relied on the skill level of the researcher.
Pollen grains were identified using comparisons from reference slides and online images, which
may be quite accurate. Volumes were estimated by taking the average of the principle diameters
of five pollen grains from each sample.
In the field experimental design, the paper makes reference to the experimental field being
surrounded by fields that had been planted with maize and soybeans, but gives no date as to
when planting occurred. This may be critical information because if the fields were planted
within a few days of the study, fugitive dust from those fields may have contributed to the
excessive concentrations found in the results. A history of surrounding land use including time of
planting should have been included in the paper.
On July 12, 2010, half of the experimental field was planted with commercial pretreated maize
using talc in the seed box at the recommended rates. The other half of the field was planted with
harvested seed from the year before. The paper does not elaborate on this seed, which may
introduce a source of error. More description about this seed should be included in the paper.
The planting equipment used is adequately described and the recommended rates of application
and procedures are clearly explained. The exhaust system was cleaned between plantings with
compressed air only. The results indicate contamination from inadequate cleaning between
plantings. This cleaning method was not appropriate and should have included a more thorough
method for pesticide residue removal.
The description of anther collection was part of the vacuum system cleaning section but should
have been a new paragraph. It was difficult to understand where these anthers were coming from
when the field had just been planted. It is assumed that they are harvested at a later date. Once
again, accurate chronological order and description is lacking.
The sampling done by the researchers in 2011 is unrelated to the previous results and should be a
separate paper. Although the information collected is very supportive of their research, it seems
out of place and should be included separately of in the discussion. It implies that the experiment
continued the following year, which it did not.
Controls were not specifically mentioned in the materials and methods section, however, the
study does incorporate several controls. The agricultural field is a control that includes both
treated and untreated seed. The presence of atrazine and metolachlor found in all bees indicates
that these bees were foraging near agricultural fields because these herbicides are commonly
applied during maize planting. In fields far away, dandelions had non detect levels of pesticides.
In evaluating this paper, there are several unanswered questions that should be addressed and
several recommendations that can be made.
Critical findings from pollen found in stored cells and traps, suggest further study is needed to
delineate between oral toxicity and contact toxicity.
Another route of exposure for honeybees that was not studied in this paper is from ingestion of
dew droplets. Samples from dew droplets should have been taken to determine if this is a
significant source of exposure. The addition of talc and water may alter neonicotinoid chemistry,
which may affect oral toxicity. This was not discussed in the paper.
The consistent use of units would improve readability and comparability.
An explanation as to why nectar did not contain neonicotinoids when the pollen did in the 2011
The 2010 and 2011 results should be separate papers because the experiment is not the same for
each year. Different methods were used at different locations so direct comparisons cannot be
made between results.
5.0 Concluding Discussion
Honeybees provide valuable ecosystem goods and services to the world in the form of
pollination and honey production. They are the best pollinators in the world and the world food
supply is dependent on them. CCD is occurring in countries that use pesticides and there is
overwhelming evidence that neonicotinoids are to blame. Countries that have taken a
precautionary approach have reduced the risk by removing the hazard and banning these
pesticides. As a result, these countries are experiencing the comeback of the honeybee.
Canada’s reaction to this problem has been slow compared to that of Europe, which immediately
banned these products upon evidence that linked them to extremely high bee mortality. Europe
has embraced the precautionary approach and has put the burden of proof back on the
manufacturer to prove they are safe. The approach in the US is exactly the opposite and the
burden of proof is on the individual to prove that harm was done. Canada’s approach falls in
between and their lack of an outward ban indicates pressure from competing interests. However,
this research supports Canada’s efforts to understand CCD and is vital to policymakers in
Canada and the US.
Although the structure and readability of this paper was inconsistent, what truly matters is the
science. The findings in this paper are important to honeybee health and they advance the
understanding of CCD. It is impossible to ignore. Also, they open the door for future studies that
delineate between oral and contact toxicity. This may have implications for the prohibition of
application during dry windy conditions or rainfall.
Determining the cause of CCD requires investigation on the various ways honeybees are exposed
to hazards. In order to assess these hazards; toxicity, foraging behaviour, and normal farming
practices were studied. Although the toxicity of neonicotinoids on individual honeybees is well
studied and documented, it was unclear that when use as directed why so many honeybees were
dying. This paper provides a better understanding of how farming practices are contributing to
honeybee deaths and CCD.
The causes of CCD are clear, however, the issue is highly politicized and the findings in research
often compete with the self-interest of multinational pesticide companies. Maintaining scientific
integrity and reporting transparent and reproducible results is not only sound science, but ethical
science, both of which have been achieved in this paper.
Anastassides M, L. S. (2003). Fast easy mulitresidue method employing acetonitrile partitioning
and 'dispersive solid phase extraction' for determination of pesticide residue in produce. AOAC
Int , 86, 412-431.
Cutler GC, S.-D. (2007). Exposure to clothiandin seed-treated canola has no long-term impact on
honey bees. J Econ Entomol , 100, 765-772.
Gauthier, M. (2010). State of the art on insect acetylcholinereceptor function in learning and
memory. Adv. Exp. Med. Biol. , 683, 97-115.
Health Canada. (2013, 10 28). Evaluation of Canadian Bee Mortaliities in 2013 Related to
Neonicontinoids. Retrieved 11 25, 2014 from http://www.hc-sc.gc.ca/cps-spc/pubs/pest/_fact-
Helman, S. (2013, 06 23). Retrieved 11 27, 2014 from www.bostonglobe.com:
Horan, L. (2010, 29 2014). Pecticide Action Network. (G. Truth, Producer) Retrieved 11 04,
2014 from http://www.panna.org/blog/neonics-not-much-help-farmers
Jeschke P., N. R. (2008). Neonicitinoids- fom zero to hero in insecticide chemistry. Pest
Management Science , 64, 1084-1098.
Krupke, C. e. (2012, 01 03). www.plosone.org. Retrieved 11 24, 2014 from
Lu C., e. a. (2012). In situ replication of honeybee colony collapse disorder. Bulletin of
Insectology , 65 (5), 99-106.
Schwartz, A. (2010, 12 15). Retrieved 11 04, 2014 from Fast Company:
Sgolastra, F. e. (2012). Effects of neonicotinoid dust from maize seed-dressing on honey bees.
Bulletin of Insectology , 65 (2), 273-280.
Williamson S.M., W. G. (2013). Exposure to muptiplecholinergic pesticides impairs olfactory
learning and memory in honey bees. Journal of Experimental Biology , 216, 1799-1807.