Another bungled correlation study on bee health dangers of neonics


What would you say about a government study claiming to show a connection between coffee and cancer — but it didn’t take into account that most of its study subjects were chain-smoking uranium miners who preferred to sunbathe without sunscreen?

You’d probably think it wasn’t very credible; or maybe it was a story on the satirical news site The Onion.

But that’s little different from a recent report issued by the usually reliable Food and Environment Research Agency (FERA) in Britain. The study, “Evidence for pollinator cost and farming benefits of neonicotinoid seed coatings on oilseed rape,” published last summer in Scientific Reports, concluded that coating oilseed rape crop seeds with neonicotinoids (aka neonics), used as a pesticide, resulted in excessive bee colony losses—what the study called a “pollinator cost”.

Here, we combine large-scale pesticide usage and yield observations from oilseed rape with those detailing honey bee colony losses over an 11-year period, and reveal a correlation between honey bee colony losses and national-scale imidacloprid (a neonicotinoid) usage patterns across England and Wales.

First off, note that this was a correlation study, not a determination of cause and effect. Many studies note correlation “links” but further exploration usually confirms that other factors, often not measured or understood, caused the problem. We know for example that the rise in numbers of people eating organic food is highly correlated with increases in obesity and autism. In other words, be cautious about putting too much weight on correlation studies.

But there are other more fundamental concerns about this report. Its conclusions are strikingly at odds with real world experience in which we have extensive data accumulated over many years. In western Canada, where 100 percent of the canola (essentially equivalent to oilseed rape) seeds are treated with neonics, bees are thriving.

And more to the point, the association study conflicts directly with large-scale field studies, which have found that bees foraging on neonic-treated crops suffer no observable adverse effects. FERA admitted as much, acknowledging in the report that no study based on real-world experience has ever been able to link the use of neonics with bee declines.

Not surprisingly, this correlation study was widely reported. “Pesticides linked to bee decline for first time in a countrywide field study,” the Guardian headlined. The headlines were turned into scare memes and promoted heavily by activists who have been campaigning against neonics and celebrate each new study that contains even a sliver of bad news about bees as a plus for its cause. Their hyperfocus on this report is understandable, as they’ve had a rough year, as recent study after study has challenged the dogma that neonics are a primary driver of bee health problems. This FERA study, they claimed for the umpteenth time, was a game changer.

Except it isn’t.

Note that this wasn’t based on field research as some activists sites claim. It was an epidemiological review that simply crunched numbers to try to find a correlation. But what the number crunchers left out is mind-boggling. One has to wade more than halfway into the text to uncover this all-important disclaimer:

Our data were not derived from a controlled experiment and so may be influenced by confounding factors not accounted for in our models, such as honey bee pests. However, controlled studies have severe logistical constraints because the required manipulations focus at the landscape scale.

That’s a starting statement, bordering on a confession that renders this research all but useless. Essentially what they are saying boils down to this: a rigorous study would have been really hard to do, so we just did what was easiest, and ignored the full picture. That’s confidence-building science?

No controls for bee pests

To translate what these scientists actually studied: they did not factor in what the European Commission (and United States government researchers) maintain are the main drivers of bee deaths: Varroa mites and the diseases it helps enable. The researchers allude to this:

Slide1Whilst the ectoparasitic mite Varroa destructor and its associated viruses have been shown to cause honey bee colony losses, no records of these or any other honey bee pests and diseases were gathered systematically for the period of interest.

In other words: we didn’t have the information about Varroa and the myriad other bee diseases readily available, so we just decided to pretend these problems — the most important ones — don’t exist. But British beekeepers can’t afford to pretend; they know bee declines are much more likely to come from Varroa infestations hitting local farms. During the period explored in the FERA study, the British Beekeepers Association explained the seriousness of the challenge they faced from this menace:

Colony losses due to varroa infestation have increased since 2001. This is a result of the mites developing resistance to available pyrethroid varroacides and the limited alternative treatments.

The correlation study also appears to be an attempt to support the narrative of a ‘crisis’ that doesn’t exist. Bee health is an ongoing, challenging issue. But in overall numbers, bees aren’t declining; not worldwide, not in Europe, not in Canada, not in the United States. Wherever you look, honey bee hive numbers have been stable or growing since neonics came into widespread use. That’s not to say there are not increased stressors on bees and legitimate concerns about how commercial bee operations manage their hives. But those are different, although related issues.

There are some worthwhile findings from the report that comport with real world observations — and common sense — although these elements have largely been ignored by the media. The study showed that neonic seed treatments work in fighting off dangerous bugs to boost yield and profitability while reducing dependence on insecticide sprays:

We combined pesticide usage and yield observations from over 76,000 ha of OSR over five replicate years to provide the first evidence that farmers who used imidacloprid seed coatings consistently reduced the number of foliar insecticide sprays used in the autumn, but not during flowering (Fig. 4). Our observations fit well with the known efficacy of such seed coatings against cabbage stem flea beetle (up to four to five weeks) and peach potato aphid (up to 10 weeks), and demonstrate for the first time that farmers do alter their foliar pesticide application practices as a result of using imidacloprid seed coatings.

By selecting this more advanced pesticide—the one that activists love to hate—you end up using less pesticides overall. Are advocacy groups cynically ignoring the well-accepted fact that neonics decrease the overall use of pesticides? Are they, in effect, advocating for increased pesticide use?

From a scientific standpoint, the FERA’s conclusion that neonics forestall crop damage isn’t exactly groundbreaking. It’s common knowledge amongst farmers, which is one reason a number of European countries have requested derogations from the European Union’s ban on neonics instituted almost two years ago. In July, Britain claimed its exemption after crops began to be overwhelmed with flea beetles. Finland, Romania, Bulgaria, Estonia and Denmark have done likewise.

So, its unfortunate that while elements of the study might have contributed to a more nuanced understanding of the debate on neonics, it was the shoddy bee-decline correlation that grabbed the headlines, supplying fodder for the anti-pesticide campaigners.

Jon Entine, executive director of the Genetic Literacy Project, is a Senior Fellow at the World Food Center‘s Institute for Food and Agricultural Literacy, University of California-Davis. Follow @JonEntine on Twitter

  • Dan Boersma

    Very timely and important info. I would like to share, however there are so many grammatical errors, it seriously detracts from the work. A good editor would be of great use here

  • Matt Shardlow

    1) Of course most scientific results are evidence and not proof,
    correlations in particular need caution – do you have any alternative explanation for the correlation?

    2) You say that “no study based on real-world experience has ever been able to link the use of neonics with bee declines”. On a tight defintion of large scale real-world data, this is the first such study, so it is interesting that it has found a correlation.

    3) You state that the study found that neonics “boost yield and profitability”, this is not true, for imidacloprid there were 3 years with no significant difference from 0, 2 yrs with a significant positive effect and one year with a significant negative effect. This lack of any consistent benefit to yields from neonicotinoid seed treatments has been repeatedly observed. A finding supported by the results of this recent study by the AHDB – “There was no statistically significant effect of treatment on crop yield.”

    3) It is not at all clear that neonics “decrease the overall use of pesticides”. The use of neonicotinoid seed treatments in this study resulted in a 14% (+/- 11.3% SD) reduction in the use of insecticide sprays on the crop (a reduction that may be statistically significant) and a 92.9% (+/-13.7% SD) increase in the number of insecticide treatments on the crop (which would clearly be significant).

    • Brion

      Do you have real world experience. I am an agronomist and seed treatments do play a huge part in the successful germination of crops. Instead of talking, why don’t you conduct your own experiment and find out for yourself. Trust me, if we had to rely only on organics, we would be a starving nation and you think food prices are high now…

      • crush davis

        Reading what he wrote, it looks like he has plenty of real-world experience talking trash about concepts he doesn’t support.

    • crush davis

      How do you know that the 92.9% (+/- 13.7% SD) “would clearly be significant?” I don’t see any F or t statistics to indicate that. Nor do I see any charts that show how the means or medians of different treatments compare. Where is your inferential statistical evidence? btw, how eager would you be to accept a correlation for some relationship you wanted to see? My guess would be, “Very.”

      • Matt Shardlow

        In correspondence with Giles Budge the lead author of the paper he confirmed that the data “did show a clear ‘statistically significant’ decrease in the total number of
        foliar insecticide applications” for imidacloprid use. This is very similar to the 14% calculation above. Adding in the seed treatments which are either one or two insecticide treatments gives the second figure. I don’t have the raw data, Giles was not willing/able to share that, so the figures are based on the annual averages. See Fig 4 of the paper. If a 14% difference is a significant or nearly significant difference then it is difficult to see how a 92.9% difference with a similar SD would not be significant.
        In this context we are considering significant differences, not correlations.

        • Brion

          Correlations are stronger evidence.

  • Jon, in so many ways this story is distorted and slanted.
    1. “In overall numbers bees aren’t declining”
    The reason this story exists is because bees are declining.
    Beginning in October 2006 bee keepers began reporting losses of 30% to 90%.
    These losses continue to date according to the USDA Agricultural Research Service.
    “CCD has resulted in higher costs for pollination and could threaten the pollination industry if it becomes more widespread,” the USDA ARS states in their website.

    2.”The Shoddy bee-decline correlation” is your opinion and not the facts:
    In January 2012 Purdue University scientists published results of their two year study showing the insecticides clothianidin and thiamethoxam, commonly used to coat corn and soybeans seeds, were killing pollinating bees.

    “We know that these insecticides are highly toxic to bees: we found them in each sample of dead and dying bees,” quoted Christian Krupke, associate professor of entomology and co-author of the findings in the Purdue University News Service.

    The Purdue Study sited the neonicotinoids as compounds that can persist for months or years with plants growing in the treated soil taking up the compounds in leaf tissue and or pollen.
    In March 2012 in the journal of Science two teams of researchers, one in France and the other in Britain, published studies showing neonicotinoids have significant negative impacts on bee health and colony survival.

    And in April 2012 the Harvard School of Public Health released their study to be published in the June issue of the Bulletin of Insectology, citing new research providing evidence linking imidaloprid and bee CCD.

    The authors of this study proved bees exposed to imidaloprid in a plant’s pollen or through the high fructose corn syrup beekeepers use to feed their bees, and it resulted in CCD. Over 90 percent of conventionally grown corn in the US has been treated with neonicotinoids and it is in corn syrup.
    The governments of France, Germany, and Italy are not waiting for more studies tying bee CCD with neonicotinoids. Since 2008 there have been bans on seed treatment using neonicotinoids in all of these countries.

    • Brion

      you must work for the gov’t or are part of gov’t to reply as you did.

      • Nope…just a New England farmer who tells it like it is.

        • Brion

          You know honeybees are not native to North America??? You know there are many other insects that are pollinators… I am just telling you like it is and folks like you don’t care to hear it. But you would rather point fingers at pesticides rather than human infringement on an insects habitat. What was once a natural habitat for many living species, is now covered with blacktop and concrete. The other is increased traffic and highways and roads. Next time, check the amount of insects (all) on your grill of your vehicles.

          • Bill

            What does being native have to do with it? All pollinators are in decline.

          • Brion

            Here I will give a left wing radical sight. So many keep talking about something they really don’t know Crap about. You ever hear of the black plague? How about AIDS? We are herding these bees and we don’t think anything will ever happen to them. We all go around believing everything will be hunky dory, but as soon as something happens, you want to point fingers at the easiest target, you know why, MONEY and GREED! Someone isn’t happy or possibly losing money cause their bees are in decline and can’t figure it out, so attack the person who does have money. Gee I wish the farmer could do that, but hey they depend on the weather, and they can’t sue Mother effing Nature!


    • trrosen

      one problem there Tex. HFCS is a refined product and contains only Fructose and Glucose. Neonicotinoids will not make it through refining.