Risk-benefit analysis of Bayer’s imidacloprid

Bayer’s imidacloprid is one of the most widely used insecticides and belongs to a group of pesticides known as neonics (or neonicitinoids), which target insects’ nervous systems. Imidacloprid binds more strongly to neuron receptors in insects than in mammals, and is therefore considered to be much less toxic to mammals, compared to insects. That’s why I’m writing this as a follow up to the last post on the 1700 pet deaths that have been linked to usage of the Seresto flea collar in the US. (The main active ingredient in the Seresto flea collar is imidacloprid.)

I really bring up the Seresto collar issue again because I want to do some risk-benefit analysis on the use of imidacloprid, and neonics in general, as insecticides. I did this in the last post, with a rough estimate that the Seresto collar posed the risk of death at around 1 in 600 pets (0.17%) based on the reports to the Environmental Protection Agency (EPA). That risk needs to be weighed against any benefit – that the Seresto collar may help reduce fleas on cats and dogs. Of course, the collar is only one of many options for controlling fleas, including some that pose no risk, like using a flea comb or vacuuming regularly.

Bad as this looks for the Serento flea collar, I believe that the use of imidacloprid and other neonics in agriculture is virtually impossible to justify in most cases. Here’s a summary of my concerns, which I will expand upon:

1. Imidacloprid is much more toxic to insects and aquatic life than to mammals.
2. Neonics are largely used by default (e.g., as a seed coating) rather than in response to an insect attack.
3. Neonics, used in this way, offer little benefit to farmers

This is one of the most critical issues that we face today and I urge you to give it some thought.

Imidacloprid is much more toxic to insects and aquatic life than to mammals

I used to just trust that imidacloprid (Advantage II) was safe and used it on my cat as a flea treatment – in other words I assumed that the risk-benefit ratio was all good. Now, knowing more, I would not use it on a pet unless it was a last resort in a life-threatening situation (and this is extremely unlikely to happen). Bearing that in mind, consider this fact again: imidacloprid is way more toxic to insects and aquatic life than it is to cats and dogs. This is an accepted fact. There’s a huge collection of evidence, so I’ll just get to some of the latest findings that you may not have heard about.

Imidacloprid is highly toxic to bees and other insects

You are probably aware by now that imidacloprid and the other neonics are toxic to bees. Bayer does its best to discredit scientific studies and delay action in an ongoing dance (battle) with researchers who rely on public (taxpayer’s) money to investigate the impact of imidacloprid. It’s a dark story of corporate propaganda and manipulation that I’m planning to cover in a future post. As far as the science goes, it boils down to studying the levels of imidacloprid that pollinators are exposed to in the field and how this impacts their behavior.

I’ve written about it several times before so I’ll just refer you to those posts:

How do neonicotinoids affect bees? – That post looks at the basics of field exposure levels versus the lethal dose

Evidence that neonicotinoids does not kill bees – An examination of data that Bayer uses to defend imidacloprid; insights into the mechanisms of Bayer’s propaganda machinery.

Subtle effects of neonics on bee health – Critical data on how exposure to sub-lethal levels of neonics can drastically reduce bee births and overall hive health.

Since then, new data keep coming out and Bayer does its best to ignore, discredit or refute them.

New study on the impact of imidacloprid on squash bees

Susan Willis Chan, a researcher at the University of Guelph, just published a study (Nature, 2021) that demonstrated a severe reproduction decline in ground nesting bees when exposed to squash crops treated with imidacloprid. Squash plants, which depend heavily on hoary squash bees for pollination, are often treated with Bayer’s product Admire, which is basically concentrated imidacloprid. Squash flowers are the sole pollen hosts for female squash bees – it’s a very tight relationship. 

Chan commented on her study: “The majority of bees in the world are solitary, ground nesting-bees and the research fills in a knowledge gap for them.”

Bayer responded to it by saying that it requires “significant additional effort to determine the relevance of the findings under realistic conditions” and also somehow managed to claim that “the bee populations were not affected.” 

Er… what? First off, the study applied Bayer’s Admire product following Ontario government guidelines. Second, take a look at the data for yourself (compare the red bars for Admire versus Control – there’s a 10-fold difference in bee births):

Notice that the outcome was reduced births – this mirrors the data discussed in my post on the subtle effects of neonics on bee health. To discuss the impact of a pesticide purely on the lethal concentration is very basic, as a lot of bad things can happen before something kills you. Let’s take a quick look at the product that was used in the Guelph study – Bayer’s Admire.

Bayer’s Admire product: 43% imidacloprid

Bayer’s imidacloprid is sold in many formats, from the Seresto flea collar to seed coatings (see later in this post) and in almost pure form as Admire insecticide (43% imidacloprid). A casual shopper can buy Admire by the gallon online. In the squash bee study above, researchers applied Admire only at the time of seed planting (18 mL/100 m row), so the 1 gallon bottle pictured below provides 210 times that amount. You can imagine that farmers may sometimes be tempted to apply more, or to spray plants multiple times as they grow.

How much safety information does Bayer provide about imidacloprid?

The safety data sheet (SDS) that comes with Bayer’s Admire fails to list warnings that are available on imidacloprid SDSs from other sources. This is a serious breach of legal safety obligations as SDSs are relied on as objective sources of information.  Basically, the Admire (imidacloprid) SDS doesn’t list the acute and chronic aquatic toxicity. In contrast to Bayer’s SDS, every other SDS for imidacloprid that I could find lists the aquatic toxicity – I’ve shown two of them below, from Cayman Chemicals and PubChem (the National Library of Medicine). Of course they do – it’s required by law.

Note that the acute (short-term) and chronic (long-term) toxicity is classed as Category 1. This is the highest level of toxicity. How can Bayer possibly get away with not listing this on the Admire SDS?

Let’s move on to looking at data on aquatic toxicity.

Imidacloprid is highly toxic to aquatic life

As you can imagine, imidacloprid is most toxic to smaller aquatic creatures such as tiny shrimp and midges than to larger creatures like fish. The impact on smaller creatures can be even more dangerous as it can go unnoticed and yet impacts the entire food chain, up to fish and birds. But recent data suggests that even fish such as rainbow trout may not be immune to direct impact.

Toxicity of imidacloprid to rainbow trout

Bayer’s SDS for Admire cites data on toxicity of imidacloprid on adult rainbow trout – the lethal concentration for 50% of the fish (LC₅₀) at 96 hours is 211 mg/L. Other SDSs have included a different study on young rainbow trout (fry) which took quite a bit of digging around to find – here’s a 1992 report from the EPA that discusses it. The study found “significant difference from control” in three parameters (survival, swimming and weight) at even the lowest level studied – 1.2 mg/L.

Because there was an impact at the lowest concentration studied, the EPA concluded that it was not possible to set a NOEC level (NOEC is the no observed effect concentration). This number should have been reported on the Bayer SDS and a follow-up study should have been carried out to determine how much lower the concentration needs to be before there is no observed effect. Instead the 30 year old report seems to have been buried.

Imidacloprid accumulates in the blood of rainbow trout

I want to discuss a 2018 paper titled “Toxicokinetics of the neonicotinoid insecticide imidacloprid in rainbow trout (Oncorhynchus mykiss).” The researchers state that imidacloprid concentrations in water systems are normally at least 100 times lower than the level (1.2 mg/L) that causes problems with fry, “with the exception of extreme circumstances.” They reference a 2015 review that provides a nice list of 10 studies on imidacloprid in water systems and actually 6 out of the 10 studies found imidacloprid concentrations between 0.014 and 0.320 mg/L. So, concentrations were actually above that threshold (100-fold lower than 1.2 mg/L) in 60% of the environmental studies, and reached peaks that were only 4 times below the level that impacted rainbow trout fry.

But even if we were to consider this as a safe margin for rainbow trout fry, the authors of the 2018 paper uncovered some new findings that did cause them concern – chronic persistence in trout. Bayer had claimed that persistence (or bioaccumulation) does not occur with imidacloprid, making it safer than older pesticides like DDT. However, the rainbow trout study found that:

The results suggest some similarities in handling of imidacloprid by trout and mammals (e.g., rapid distribution) as well as striking differences (e.g., an apparent absence of biotransformation in trout). Importantly, branchial clearance of imidacloprid by trout was much lower than would have been predicted based on non-specific partitioning to blood constituents. This was likely due to specific binding of imidacloprid to plasma proteins and/or hemoglobin. This latter finding is significant insofar as it suggests that imidacloprid would tend to accumulate in trout in continuous aqueous exposure.

So that’s not good, and we weren’t even expecting the possibility of a direct impact on fish – more on the insects, shrimp and other small creatures that provide a foundation for aquatic habitat food chains. Then again, we weren’t expecting an impact of imidacloprid on cats and dogs either. The paper concludes:

Given the widespread use of imidacloprid, as well as its documented presence and relative persistence in surface waters, additional research on uptake and accumulation in fish appears to be warranted.

Let’s take a look at the impact of some of the smaller aquatic creatures.

Toxicity of imidacloprid to small aquatic organisms

As mentioned, the smaller aquatic species are even more sensitive to imidacloprid than rainbow trout and they are important parts of both aquatic and avian food chains. I’m going to have to switch to micrograms (µg) now as these smaller species are so much more sensitive to imidacloprid. So when I mentioned above that 6 out of the 10 environmental studies found imidacloprid concentrations between 0.014 and 0.320 mg/L, this is the same as saying between 14 and 320 µg/L. Also, µg/L is same as parts per billion (ppb).

Bayer’s SDS for imidacloprid (Admire) mentions that the LC₅₀ for the harlequin fly (Chironomus riparius) is 55 µg/L but only lists one other insect species. Check out the table below from a 2014 report from the Netherlands government that summarized studies (which they screened for reliability) on toxicity of imidacloprid to various aquatic species. The left side of the table lists the levels (LC₅₀) at which imidacloprid is acutely toxic – i.e., causes death quickly. You can see that several crustacean and insect species will quickly die when exposed to imidacloprid levels as low as 1 µg/L.

Then on the right side of the table you’ll find the NOEC levels for various species and the concentration that is considered to be safe (no observable effect) is well below 1 µg/L for more than a third of the crustaceans and insects studied, with two having values around 0.02 and 0.03 µg/L. That’s 40,000-60,000 times lower than the concentration that impacts rainbow trout fry (which was already concerning).

How much imidacloprid in our water systems is considered acceptable?

Based on their findings (the table above and other factors) the Dutch government recommended that their current MAC-EQS of 0.2 μg/L should be maintained and proposed a new AA-EQS of 0.008 μg/L.  (MAC = maximum acceptable concentration; EQS = environmental quality standard; AA = annual average).

A Water Quality Criteria Report for Imidacloprid that was published by University of California (Davis) researchers in 2018 provided guidelines to the California Central Coast Regional Water Quality Control Board. The researchers found that the lowest acute imidacloprid toxicity value (LC₅₀) for the ostracod, Cypretta seuratti, at 0.07 μg/L. The Californian researchers proposed safety limits that in the same ballpark as the Dutch numbers: 0.07 μg/L as the maximum acceptable (acute) concentration and 0.014 μg /L as the long-term (chronic) average limit.

Remember that 60% of the environmental studies that I mentioned above found imidacloprid concentrations between 14 and 320 µg/L. Even taking the highest of the 4 limits above – the Dutch maximum acceptable concentration of 0.2 µg/L, the environmental measurements are between 70 and 1,600 times higher than this level. The UC Davis report listed a few imidacloprid measurements by the California water quality board. It is shocking how seldom imidacloprid has been quantified, considering that the numbers are not good: 9.14 µg/L imidacloprid was measured in the Santa Barbara County water supply – 45 times greater than the Dutch limit.

Bayer’s directions for Admire allow unsafe levels of imidacloprid in water systems

Bayer’s directions for using Admire on various crops allows for the use up to 0.5 lb of product per acre of land (an acre is about 60% of a soccer pitch). 43% of that product is pure imidacloprid, so that’s roughly 195 grams of imidacloprid. If you poured that same amount of Admire into an Olympic-sized swimming pool (not an unrealistic amount of water for an aquifer under one acre of land) this works out at about 78 µg/L imidacloprid – 390 times greater than the Dutch limit and 1,114 times higher than the acute limit set by the University of California researchers.

Let’s now look at the usage and possible benefits of imidacloprid so that we can weight it against the risk.

Neonics are largely used by default rather than in response to an insect attack.

As mentioned above, Bayer’s directions for using Admire recommend the application of up to 0.5 lb of product per acre of land. This is often applied at the same time as planting seeds – before the plant even starts growing. In other words, the pesticide is applied at concentrations that pose an unacceptable risk as a default course of action rather than a last resort against crop damage. This is true not just for Bayer’s liquid Admire product, but also for the majority of seeds (for both agriculture and gardening) sold in many countries outside the EU.

Seeds coated with imidacloprid and other neonics

In case you didn’t know, a large proportion of the corn and soy grown in the US is from seed that’s coated with imidacloprid and/or other neonics. This is also true for Canada, much of South America and Asia, and many countries across the world. The use as a seed coating very much fits the definition of use by default.

A single treated oilseed rape seed (e.g., canola) is typically treated with approximately 35 mg neonicotinoids and a corn seed with 1.2 mg – that’s a lot of pesticide on one seed. The impact of imidacloprid on birds is within this scale – for example, in mallard duck and quail reproductive studies, effects on eggshell thickness were observed at concentrations of 61 mg/kg-diet. A few imidacloprid-coated rape or corn seeds foraged from a field would meet that level. 

In a sad study published this month by Argentinian researchers, grayish baywings fed with millet seeds treated with imidacloprid all died within three to five days of exposure. The authors calculated that if the birds’ diet comprised over 7% imidacloprid-treated seeds, they would likely die. I won’t go on, as you’ve probably heard enough – but I could, because there’s so much data out there. A recent article from National Geographic reports on the impact of neonics on birds and even rabbits and deer.

Consider that only about 5% of the neonics in seed coatings actually make it into the plant – the remaining 95% remains in the soil or is washed into the groundwater and streams, etc.

Neonics, used by default, offer little benefit to farmers

Two scientists working for the US Dept. of Agriculture (USDA) studied the impact of neonic seed coatings on soybean crop protection and yield. Here’s one of the main findings:

The laboratory trial revealed that all bioactivity of the seed treatments against soybean aphids was gone within 46 days after planting, prior to aphid populations damaging the crop.

In other words, the seed coating was no longer effective by the time that aphids might want to eat the plants. Even worse, the neonics that were tested (imidacloprid and thiamethoxam) actually reduced populations of the natural predators that would normally help control the aphid.

The other main finding: The resarchers found “no yield benefits of insecticidal seed treatments over the 2 years of the study” and concluded:

This work confirms that insecticidal seed treatments offer little benefit to soybean producers of the Northern Great Plains and adds to the discussion by suggesting that generalist predators are adversely affected by the insecticides.

If you want more macro data on the “benefit” of neonics, consider that corn and sunflower yields were unchanged after imidacloprid was banned in France.

Farmers and consumers in general are often not even aware that the seeds they’ve purchased are coated with neonics. We have this idea that the farmers might need these seed coatings and select them based on experience. The reality is that the agrichemical companies have bundled their wares into one package – glyphosate-resistant seeds that are also coated with neonics and perhaps fungicides too – and there’s just an assumption that it must be beneficial.

Even those who go out and buy a bottle of Bayer’s Admire could be completely unaware of its harmfulness as Bayer has neglected to place the required pictogram and warnings about aquatic toxicity on the product.

The use of imidacloprid in agriculture is not justified

Shocking as the flea collar story might be, the risk/benefit analysis for the use of imidacloprid in agriculture is much more extreme. The way imidacloprid is used in agriculture would be analogous to me taking antibiotics all the time, just in case I get an infection. Nobody does this because we know that this is not a good risk/benefit scenario – the constant use of antibiotics would mess up our own gut microflora, and 99.99% of the time they would not even be needed. In fact, the antibiotics would become less effective over time as pathogens develop resistance after being exposed to them for so long.

This brute-force approach to agriculture has the same downsides. As mentioned above, neonics are so widely toxic that they’ll kill off beneficial insects, like the natural predator of the aphid, or the bees that pollinate many of our plants. Their use by default has resulted in insects developing resistance to neonics – just as the bacteria in our guts would develop resistance to antibiotics, if we took them all the time. Their widespread use has actually led to changes in the plants themselves, as a study on rice has shown.

Imidacloprid increases susceptibility of rice to the brown planthopper

In a 2012 paper, researchers examined one hypothesis that rice has become susceptible to the brown planthopper (PBH) because of the overuse of neonics.

In Southeast Asia and China, the resurgence of the BPH is thought to have been caused by the use of chemical pesticides. The side effects of pesticides include the destruction of natural enemies, stimulation of reproduction of male and female adult BPHs, and pesticide-induced susceptibility of rice to the BPH.

The conclusion was that the application of imidacloprid-induced stress-response genes (that’s the rice attempting to detox after exposure to imidacloprid) and this increases susceptibility of rice to the BPH. It also reduces abundance of the PBH’s natural enemies, as the USDA group had found with the soybean aphid.

The reduction of resistance of rice to insect pests due to physiological and biochemical changes in rice plants caused by pesticide applications is known as pesticide-induced susceptibility

The application of imidacloprid also reduced the rate of photosynthesis in the plants in addition to increasing the level of damage caused by the BPH.

Why then are neonics so widely used?

Of course there are some cases when higher yields can be achieved through the use of neonics. But these increments in yields are small, even when they do exist. As mentioned above, neonics are often used unknowingly – farmers, landscapers and casual gardeners buy seeds that they don’t even know are treated with neonics. Others trust that they must be good and surely there’s lots of data out there? As Tom Philpott describes in his 2020 book, Perilous Bounty: The Looming Collapse of American Farming and How We Can Prevent It, Bayer and the other agrichemical giants have been recently acquiring the major agriculture data analytics companies. The goal: complete control, tying up not just the seeds and the chemicals but also the data that informs their use. 

Is this risk-benefit scenario for imidacloprid acceptable?

Of course it’s not! Anyone who looks at the various numbers above would have to agree with that. This risk is way too high – particularly to pollinators and other insects and to aquatic habitats, but also to birds and other creatures. The demonstration that imidacloprid can accumulate in the bloodstream of rainbow trout is a new finding and suggests that the longer we continue to use neonics the greater the damage will be. Bayer had always claimed that imidacloprid does not bioaccumulate.

You would imagine that the considerable downside is balanced by clear benefits of using neonics. I’m sure there must be some data somewhere but I’ve yet to see a publication demonstrating a positive impact of the application of imidacloprid to seeds on crop yield. There is, however, quite a bit of data showing that neonics offer no benefit to yield and can actually cause insect problems such as pesticide-induced susceptibility and the destruction of natural pest enemies (“beneficial insects”).

As I reported in 2018, the EU voted to make its ban on neonics permanent (although the UK flirted with the idea of reintroducing them this year, post-Brexit). The US, Canada, and many countries across South America, Asia, Africa, and the rest of the world still use neonics to the extent that if you buy products or veggies that are not pesticide-free or organic, they were probably treated with neonics (especially in the US). This is why I’m disappointed when I find that a product that would otherwise rate well, ethically, is made from conventional soy, corn, or sunflower oil, etc.

I get irritated by organizations that exaggerate a situation to help shock people into supporting their cause. As a scientist with no ulterior motives I’ve done my best to represent the situation as objectively as possible here. My conclusion is that it is time to act to restrict the use of neonics. We don’t need any more research or delay tactics – we’ve already have more than three decades of this.