ResearchBlogging.orgA commonly used insecticide, and possibly an increasingly widely used form of that pesticide, could be a causal factor in bee colony collapse. It is not 100% certain that this pesticide’s effects can be counted as one of the causes this problem, but there is a very good chance that neonicotinoids can cause a drop in hive population, and thus, should be examined to see if there is a relationship in some cases. From the paper’s abstract:

Nonlethal exposure of honey bees to thiamethoxam (neonicotinoid systemic pesticide) causes high mortality due to homing failure at levels that could put a colony at risk of collapse. Simulated exposure events on free-ranging foragers labeled with an RFID tag suggest that homing is impaired by thiamethoxam intoxication. These experiments offer new insights into the consequences of common neonicotinoid pesticides used worldwide.

CCD (Colony Collapse Disorder) was identified as a phenomenon in which colonies of honey bees (Apis mellifera) would suddenly disappear from their own hive. Suggested causes have included pathogens, parasites, habitat loss, and even pesticides. So far studies have failed to make any clear links.

Pesticides have always been on the table as a possible cause because these days so many bees forage on industrial grown grasses (corn, sunflower, etc. etc.) which are periodically dusted with insect killers. In particular, Neonicotinoid pesticides, which are used to protect crops against sap-sucking insects (such as aphids) are associated with loss of functional behaviors such as navigational skills. Bees that are exposed to non-lethal doses of Neonicotinoid pesticides forage abnormally, have “olfactory memory” problems, are easily disoriented and become poor learners. Essentially, a dose of this sort of pesticide may make it impossible for a bee to return to the nest. If an entire colony of bees is infected by this chemical, by direct exposure as well as indirect exposure (among bees back in the hive after foraging bouts) the entire colony cold just fly off and not come back over a fairly short period of time.

In the present study, the research team “tested the hypothesis that a sublethal exposure to a neonicotinoid indirectly increases hive death rate through homing failure in foraging honey bees,” focusing on the recently marketed form “thiamethoxam.” They exposed bees to this chemical then measured their ability to find their way home, using tagged bees. Then, they modeled the effects of the combination this problem of getting lost with other problems bees have while foraging, like getting eaten and such on colony viability. For the later, they simply modeled honey bee population dynamics with the “I got lost because I ate thiamethoxam” factor added in.

The bees were tracked with a very tiny device called a RFID device, as in this photograph from the Science paper:


The researchers tagged 653 individual bees over four separate treatments, in an area of France.

To simulate intoxication events, foragers received a field-realistic, sublethal dose of thiamtethoxam… and were released away from their colony with a microchip glued on their thorax. RFID readers placed at the hive entrance were set to detect on a continual basis tagged honey bees going through the entrance. Mortality due to post-exposure homing failure, mhf, was then derived from the proportion of non-returning foragers.

Other bees, not dosed with the poison, were also monitored for comparison.

The bees dosed with the poison were less likely to return to the hive if released a kilometer away, than those not treated. In the case of bees released at a place they had been to before, from which they could use their internal bee-ish mapping abilities to fly home, the treated bees were very slightly less likely to return to the colony. Bees that were released at random locations that they were not likely to be familiar with were more discobobulated, and while both treated and untreated bees had a hard time making it back to the colony, the difference between the two was more dramatic.
A: Bees released at a location known to them a kilometer away from the hive; B: Bees released at a random location a kilometer away from the hive. The gap in the graph is overnight.

One thing that strikes me as especailly interesting here is that many bees don’t make it back over a fairly long period of time even under normal conditions, and that some bees stay out overnight!

Once these numbers were obtained from empirical fieldwork, they were plugged into models for bee population dynamics. The result is the following graph (see caption for explanation):
Comparison of honey bee population dynamics between simulated colonies exposed to thiamethoxam (red lines) or not exposed (blue lines), following six demographic scenarios. L: queens daily laying rate (nb. of eggs per day). Exp: proportion of foragers exposed to treated crops during the day. The non-exposed colony follows either (A and D) a normal development trajectory (at L=2000), (B and E) an equilibrium dynamic (L=1800), or (C and F) a slightly declining trajectory (L=1600). Shaded areas delineate the exposure period (e.g., oilseed rape). Pairs of trajectories in exposed colonies were obtains with the lower and upper bounds of homing failure mortality (0.102 and 0.316), in order to delineate the best and worse estimates for population dynamics, respectively. Dotted lines extend the declining trajectory expected for a sustained exposure.

As you can see, colonies with bees exposed to a disorienting chemical are more likely to reach a critical level of population size. In fact, the difference between non-affected and affected colonies can be quite dramatic, and for sustained exposure, heading towards the point of no return is well within the range of possibility over a period of just a few months.

The researchers conclude:

Our study clearly demonstrates that exposure of foragers to non-lethal but commonly encountered concentrations of thiamethoxam can impact forager survival, with potential contributions to collapse risk. Furthermore, the extent to which exposures affect forager survival appears dependant on the landscape context and the prior knowledge of foragers about this landscape. Higher risks are observed when the homing task is more challenging. As a consequence, impact studies are likely to severely underestimate sublethal pesticide effects when they are conducted on honey bee colonies placed in the immediate proximity of treated crops. Finally, this study raises important issues concerning exposed solitary bee species, whose population dynamics are probably less resilient to forager disappearance than honey bee colonies.

Henry, M., Beguin, M., Requier, F., Rollin, O., Odoux, J., Aupinel, P., Aptel, J., Tchamitchian, S., & Decourtye, A. (2012). A Common Pesticide Decreases Foraging Success and Survival in Honey Bees Science DOI: 10.1126/science.1215039


  1. #1 Phil
    March 31, 2012

    Just a moment.

    Parasitic flies are the cause of CCD.

    Or a virus. Or a fungus.

    Which cause am I to believe? Is it a combination? Or too soon to say?

  2. #2 Gar Lipow
    March 31, 2012

    Umm I think it is pretty clear that this is a multi-causal phenomonena. But at this point it is clear as the post and the studies say that this type of pesticide is a significant contributor. Some of the other causes you mentioned also have clear and convincing evidence. For others it is too soon to say,

  3. #3 Bacopa
    March 31, 2012

    Thanks for emphasizing the importance of pollen foraging here. Everybody thinks about nectar foraging and honey production, but pollen is an important source of protein for the growing larvae. Later in the season, the bees get enough pollen incidentally from negtar foraging, but in the spring, bees want pollen as much as they want nectar. There are lots of larvae to feed and still plenty of honey left over from the winter. Pollen is the resource bottleneck and they will forage from non-nectar producing plants to get it.

  4. #4 bug_girl
    April 1, 2012

    Gotta disagree with your headline, Greg.
    Colony collapse is a complex syndrome, and this only hints at a possible cause. Phil in comment 1 is right–this is not the story.

    CCD is characterized by these factors:

    Collapsed colonies
    1) complete absence of adult bees in colonies, with few or no dead bees in or around colonies,
    2) the presence of capped brood, and
    3) the presence of food stores (both honey and bee bread) that are not robbed by other bees or typical colony pests (small hive beetles, wax moths, etc.). If robbed, the robbing is delayed by a number of days.

    Collapsing colonies
    1) an insufficient number of bees to maintain the amount of brood in the colony,
    2) the workforce is composed largely of younger adult bees,
    3) the queen is present, and
    4) the cluster is reluctant to consume food provided to them by the beekeeper.

    I posted a longer article discussing this and other new research on my blog sunday. ( )

    Likely causes of CCD remain:
    increased losses due to the invasive varroa mite (a pest of honeybees);
    new or emerging diseases such as Israeli Acute Paralysis virus and the gut parasite Nosema;
    pesticide poisoning through exposure to pesticides applied to crops or for in-hive insect or mite control;
    bee management stress;
    foraging habitat modification
    inadequate forage/poor nutrition and
    potential immune-suppressing stress on bees caused by one or a combination of factors identified above.

  5. #5 gregLaden
    April 2, 2012

    My title? Should I change ‘could cause’ to ‘could possibly cause’ .

    … Or is something that causes a condition as documented not causal if it is not a sole cause?

    or is it just that I’m not authorized to talk about ‘bugs’?

    Anyway, thanks for your excellent and insightful comments.

  6. #6 Bug Girl
    April 2, 2012

    these pesticides may cause changes in bee behavior, or even kill bees–but that is NOT the same as causing CCD.

    As I said in my earlier comment, to call something CCD several additional criteria other than “bees gone” must be met.

    Additionally, there hasn’t been any consistent correlation between use (or absence) of these pesticides. The results are tremendously variable. That makes sense in an extremely complex biological system–but at this time, there is no single cause of CCD, and I don’t think there will ever be. It’s clearly a multi-factorial issue, and the two leading issues seem to be bee parasites and diseases (virus and fungal); the pesticides used to treat those parasites show up more consistently in CCD hives than neonics. Everything else in the list above seems to be synergistic–they make things worse, but don’t start the process in isolation.

    As for the “you’re not authorized to talk about bugs” comment: *dopeslap*
    Sheesh. ;p

  7. #7 Greg Laden
    April 2, 2012

    Bug: I get that CCD is complex, that’s been known for a while (and not just “complex” as in “we are baffled so it must be complex”)

    I was careful to use the original paper’s abstract and conclusion text as direct quotes:

    “…exposure of honey bees to thiamethoxam … could put a colony at risk of collapse. … hypothesis [is htat ] neonicotinoid indirectly increases hive death rate through homing failure … exposure of foragers to … thiamethoxam can impact forager survival, with potential contributions to collapse risk..”

    I may be guilty of leaving out “possibly one of many possible causes” etc. etc. … it is also true that I did not cite other research or papers in my post. But my post was a highly focuses description of a specific published paper. I promise … I leave stuff out of blog posts all the time. Were that not true, I’d still be writing my first blog post ever because everything is connected to everything else! And there is no lie by omission here …. I’ve accurately described the paper. The paper is about specific research which stands or falls on its own. What is wrong with the research reported here, specifically? Did they use an incorrect population dynamics model? Did they replicate exposure unrealistically? Did they use the wrong kind of bees?

    It seems to me that they are saying that neonicotinoid can cause a situation in which a colony is at greater risk of collapse because its numbers go below optimal or even to some very low level which would cause a variety of stresses, leaving open the possibility of all sorts of bad things going wrong. No one is saying that Neonicotinoid explose is colony collapse disorder. How can neonicotinoid not be considered a possible “cause” while other possible “causes” are all considered only as part of a complex system? In other words, I’m sensing that neonicotinoid is playing in a different stadium here.

    Your point about a lack of correlation confuses me .. perhaps I’m dizzy from the slap upside the head, not sure … :) because if the cause is complex in a certain way … multiple factors with some subset being sufficient … then certain “causes” would not necessarily show a very good correlation.

    I get the part about CCD being a syndrome with certain criteria, and it is quite possible that neonicotinoid is an entirely different problem (which may or may not be a problem) but this study is not an examination of CCD, but rather, an examination of the effect of neonicotinoid on specific bee behaviors which are then examined in a simple model of population dynamics to see if there is a point at which colony numbers could become very low, which could then in theory be linked to other factors (not examined in this narrowly defined research project, reported here in a narrowly defined blog post) to contribut to CCD.

    I could see that there would be a problem if this chemical exposure/injestion contradicted known CCD criteria, but I don’t think that is the case for the most part. Let’s look at your criteria:

    1) complete absence of adult bees in colonies,

    Since the bees don’t come home, that partly matches. If neonicotinoid was a single cause, this would not match, but CCD is already thought to have multiple causes.

    1b)with few or no dead bees in or around colonies,

    Neonicotinoid effects (if true) would cause the affected bees to be no where near the colony (they get lost, apparently) so that fits.

    2) the presence of capped brood,

    That does not not fit.

    3) the presence of food stores (both honey and bee bread) that are not robbed by other bees or typical colony pests (small hive beetles, wax moths, etc.). If robbed, the robbing is delayed by a number of days.

    That does not not fit …. (I taek this criterion to indicate that the colony was not invaded by evil pirate bees or ants or something)

    In collapsing colonies from your comment:

    1) an insufficient number of bees to maintain the amount of brood in the colony,

    That fits nicely

    2) the workforce is composed largely of younger adult bees,

    This does not contradict, could fit, could contradict. Depends on the differences between younger and older bees.

    3) the queen is present, and

    She would be present because she did not wander off and get lost, so that fits.

    4) the cluster is reluctant to consume food provided to them by the beekeeper.

    I have no idea… Maybe the beekeeper needs to take a hint?

    I am having a hard time seeing how neonicotinoid, to the extent that it is demonstrated in this research, can’t be added to the list of causes for CCD. I do see where a very large scale correlation study could rule it out; If CCD occurs at identical rates in two identical regions with otherwise identical risk factors (those you mentioned) but where one area has neonicotinoid use and the other does not, then that would be interesting. I also think it would be difficult to come up with that comparison.

    On the other hand, it would be helpful to show that colonies affected (as indicated) by neonicotinoid caused loss of foragers are then more susceptible to other factors on the list of CCD culprits in some way additional to simply having lower numbers. (But lack of a link there does not mean much).

    Is there something I’m missing here?

    (I admit I’m always a little suspicious when industry-related causes in an industry that funds much of the research are written off in settings where we don’t have well demonstrated and fully agreed on models. Not that I’m suggesting that YOU are an industry shill!!!!)

    I wonder if we can work a discussion of CCD into the CON sessions this year, which you will be attending, right??????

  8. #8 Peter
    April 2, 2012

    So we need to start engineering bee strains resistant to Marathon? :-)

  9. #9 Lou Jost
    April 2, 2012

    There is a bright side to this, at least in areas that do not depend on bee-pollinated agriculture. Don’t forget honeybees themselves are invaders from the Old World. They can cause great harm to natural ecosystems, by outcompeting native bees and thus reducing pollination of flowers specialized for native bees. In tropical South America, where I live, Africanized honeybees are a life-threatening scourge (they regularly kill even bulls here!) and surely have a huge impact on the diversity and abundance of native bees. I manage a set of ecological reserves aimed at protecting endemic plants (many needing specialized pollinators), and I am hoping someone develops (or has developed) baits to selectively eliminate honeybees. Research into colony collapse might suggest new clever ways to eliminate them when that is desirable. If anyone knows how to selectively eliminate them over large areas, let me know. Not only my organization but many others here would love to try them out. I imagine a bit of juvenile hormone or its mimics in a sugar solution might do the trick?

  10. #10 Bug Girl
    April 2, 2012

    Your interpretation of this paper is fine, but it also needs to be interpreted in the context of the last 6 years of research on this topic.

    You have hit on part of the problem in your proposal of a large scale correlational experiment. There just is is no consistency between areas of high pesticide use, and areas where lots of hives fail and that failure is attributed to CCD. If only there were! They have managed to reproduce a major symptom of CCD, which is big news..
    But that is not *causing* CCD.
    (I am also tempted to say “It’s only a model!” but will resist Monty Python references in this context.)

    The things that do consistently show up in hives with CCD are mites, viruses, nosema, and fungus. Some pesticides–the ones used to control the mites, nosema, and fungus–also show up consistently, and have some nasty effects on bees. Those seem to be the primary triggers. My primary quibble is with the word “cause” in your title. I would say “contributes to”. If it was truly causal, I would expect that we would see it more consistently; or find more consistent patterns in the literature.

    Look at how much dissension there was in the folks interviewed in Carl Zimmers piece. It is far from settled science.

    Having said all that–and since you hinted I’m a pesticide shill :D–let me be clear that these pesticides are NOT good for bees. Personally, though, I find their effects (and potential effects–since a lot of this is still unknown) on our native bees and pollinators of far greater concern than the honey bees.

    There is no beekeeper taking care of solitary bees. They are ….solitary. They get one shot and it’s gone. Honey bee hives are provided a home, and food, and protection.
    Bumble bees and solitary bees have lost habitat, lost food plants, and are generally SOL. Add into this a poisonous nectar source, and it is not a pretty picture. :(

    I definitely think that labels for all neonics available over the counter should be changed immediately, especially in light of the finding that they may be applied (or over applied) up to 150 times the allowable agricultural dose.
    It’s also not a bad thing to limit chemicals that have begun to be used prophylactically in agriculture. In other words, farmers put these chemicals on *before* they see any pests, which is bad pest control practice. (Smart farming, though–it protects their crop investment. I can’t blame them for that, but we need to re-examine that. )

  11. #11 Greg Laden
    April 2, 2012

    All good points. I’ve modified the title and the beginning of the post to better reflect the complexity of the issue.

    It is notable, and this is the other part of the issue you mention related to which insects are affected, that honey bees are, when not in a hive, semi-invasive (but not successful in many/most US habitats). Also, of all domesticated animals, they are one of the few that forages freely if we don’t count range land as free (which it is not). I wonder what this says about who is doing what to whom? If somebody was doing something to BLM grasslands that caused a certain percentage of cattle to not return for “harvesting” payments to cattle farmers would be made. In fact, that is what happens when the Department of Interior encourages wolf or other predator populations on one hand and leases grazing land on the other; farmers are paid for their predated cattle.

  12. #12 Lou Jost
    April 2, 2012

    Bug Girl, I am glad to find that someone else cares about the native bees!!! You mentioned the effects of pesticides on the native bees; do you know if there is much data about the reduction of native bees due to competition from the much-more-efficient honeybees?

  13. #13 Phil
    April 2, 2012

    Bug girl, I hope you post on environmental websites which are jumping on this one study and blaming pesticides for everything.

  14. #14 Benjamin Raucher
    April 3, 2012

    Wow!1 The tracking of bees. Now I have seen everything.

  15. #15 Peter Borst
    April 3, 2012

    What is being overlooked is the real world experience: canola growers have been using both GM seeds and chemical treated seeds for years now, and beekeepers have been getting lucrative honey crops from the canola flowers — not dead hives!

    * * *

    Canola is grown commercially mostly on the prairies in Canada. In 2008, 16.6 million acres (6.6 million ha) were planted and the acreage is expanding. There are 52,000 canola producers. Canada is the largest single producer of canola in the world.

    Commercially grown canola is predominantly a prairie crop. It is so common that 80% of Canada’s honey crop is from canola. This amounts to 50 million lb per year of Grade No 1 white honey.

    Approximately 300,000 colonies harvest open pollinated canola. The expanding hybrid seed production industry, where farmers produce seed under contract to the seed companies, required 80,000 colonies in 2008 for pollination in southern Alberta.

    Most canola seeds are now treated with systemic insecticides such as Gaucho® (imidacloprid), Poncho® (chlothianidin) or Helix® (thiamethoxan). Although there is an expressed concern by many beekeepers around the world about the use of systemics, the experience in Canada is that we have had 10 years of large scale use on canola with no observed ill effect.

    Pollinating Hybrid Canola – the Southern Alberta Experience
    Heather Clay, Chief Executive Officer, Canadian Honey Council, Calgary, AB

  16. #16 ziekenhuisdieet
    April 3, 2012

    this is not bee-autifull.
    Cant believe that pesticides do this…
    Are there alternatives for pesticides?

  17. #17 Greg Laden
    April 4, 2012

    But wait, Peter, to be devil’s advocate for a moment, you did just state that the current Canadian canola situation is about a decade old. The colony collapse thing has been going on for about a decade. The fact that everyone is using these pesticides and the fact that there are problems with bees, together, is not evidence that there is not a link. It may not be evidence of anything at all, but it doesn’t really convince me of anything.

  18. #18 Adam K
    April 4, 2012

    Still not really convinced… Surely pesticides may play some role, but as thers pointed out, this seem to be much more complex phenomena…

  19. #19 Craig Bart
    April 5, 2012

    If there are larva/eggs sufficiently young enough the workers can use these eggs to develope a new queen. The worker bees produce a special food called Royal Jelly which causes a regular egg/larva to develop int a queen.

  20. #20 Thesius
    April 5, 2012

    Pesticides are nerve toxins. Why is this even a discussion? Oh, the PhDs are paid to defend these toxins and parrot whatever continues to support commercial viability. In fact, the scientists often play a very vital role in the toxic game of genetic roulette, they assign allowable levels of these toxins that nature, insects and humans are deemed to be able to ingest….safely. It’s the equivalent for “taking one for the team” so a few may profit at the expense of the many….or in the case of bees, the entire colony.

  21. #22 Ana
    April 6, 2012

    “Researchers have now found that repeated low-dose exposures are perfectly capable of gradually killing of whole hives of bees. In fact, 94 percent of hives whose bees had been fed the pesticide died off entirely within less than six months, according to a new paper that will be in the June issue of Bulletin of Insectology.”