Glyphosate and the Gut Microbiome: Another Bad Argument Annihilated


Glyphosate is a broad spectrum herbicide that was first introduced by the Monsanto company in the 1970s under the brand name Roundup. The already popular product grew even more popular among farmers upon the introduction of various commodity crops which were genetically engineered to resist the herbicide while it killed the surrounding weeds with which the crops would otherwise compete for water and nutrients. Glyphosate went off patent back in the year 2000, and since then many manufacturers have cashed in on its popularity [1]. Although it is of unusually low toxicity, glyphosate receives a level of scrutiny and vehemence of criticism that is disproportionate to its actual established risks [2],[3],[4]. This is attributable in part to its ubiquity in modern conventional farming, but it’s likely even more attributable to its association with Monsanto, against which a large and well-organized counter-movement has emerged [5].

Consequently, many different arguments have been formulated and circulated among this counter-movement and beyond. The purpose of this piece is address one of those arguments in particular. More specifically, on numerous occasions I have heard glyphosate critics argue that glyphosate should be opposed because it might alter the microbiome in humans. In a post on his facebook page, The Mad Virologist discussed a recently published study on the effects of glyphosate on gut microorganisms, and inspired me to unpack the microbiome argument against glyphosate and explain what’s wrong with it.


Glyphosate binds to and inhibits the action of an enzyme known as EPSP synthase, which plants need in order to make three important aromatic amino acids: phenylalanine, tyrosine, and tryptophan via what’s known as the shikimic acid pathway, which occurs in plants, bacteria, fungi, algae and some protozoan parasites [6],[7]

Image c/o Zucko et al 2010 [37].

Glyphosate does this by acting as what’s called an uncompetitive inhibitor. That means that it can only bind to the enzyme-substrate complex – the substrate being shikimate-3-phosphate in this case – and cannot bind the enzyme when the substrate is unbound [8],[9]. Upon binding to the enzyme-substrate complex, glyphosate prevents the complex from forming its product, 5-enopyruvylshikimate-3-phosphate (EPSP). Normally the complex would form EPSP by reacting with another molecule called phosphoenol pyruvate (PEP), but sufficient concentrations of glyphosate reduces the number of units of the enzyme-substrate complex available to form their product. The shikimic acid pathway doesn’t exist in us. Humans and other mammals, for example, can’t make those amino acids at all to begin with, so we get them directly from our food. Plants need those amino acids in order to grow and to make proteins, so if they are unable to synthesize them, they can’t grow, and therefore they die.

Additionally, mammals such as ourselves have lived in co-evolutionary association with myriad microorganisms whose aggregate is referred to as the microbiome. The roles of the microbiome in human health and the effects resulting from changes in its composition are active areas of scientific investigation [33].

Image: retrieved from

However, our collective knowledge of the relationship between the microbiome and human health is still in its infancy. Consequently, the topic is an easy target for exploitation by proponents of pseudoscience who would leverage it as a promotional tool for their own agendas, and/or extrapolate to claims which overstep what the current body of scientific literature actually supports [34],[35],[36].

The Gut Microbiome Argument Against Glyphosate

Keeping that in mind, the reasoning underlying the gut microbiome argument against glyphosate can be summarized as follows:

1. The makeup of a person’s gut microbiome is relevant to human health in ways which are only recently starting to be elucidated.

2. Bacteria possess the shikimic acid pathway and can use it to synthesize aromatic amino acids.

3. Glyphosate inhibits a key enzyme used in the shikimic acid pathway.

4. Therefore, glyphosate might be altering people’s microbiome in detrimental ways.

Simple enough?


Is This Argument Biologically Plausible?

However, the problem is that this argument flies in the face of one of the basic principles of microbiology: that microbes grow in the presence of abundant nutrients. As I’ve explained on several occasions when this has come up in my facebook comment sections, bacteria shouldn’t need to synthesize aromatic amino acids when they are literally bathing in them in the gut, therefore this argument against glyphosate is grasping at straws and not plausible. A recent study tested this more formally (in vivo) [10].

How did I know in advance that this was extremely unlikely to be a major issue before this research? It was because I knew that gut bacteria live in… Wait for it… the GUT!!! Where aromatic amino acids are abundant. That means they will continue to grow if the final product of a given biosynthetic pathway is supplemented to them – which is what we are doing by supplementing them with aromatic amino acids through the food we eat – even in the presence of something that inhibits that specific pathway.

This principle is the basis for experiments that allow scientists to functionally characterize which genes’ enzymes act on which substrates in a given biochemical pathway (called functional complementation analysis), and has been in common use for the last century or so as a method for ascertaining the specific steps of varous metabolic pathways, and/or the genes which code for the enzymes which catalyze each reaction [11],[12],[13],[14].

For an example of how this immersion technique has been used, consider the elucidation of the arginine synthesis pathway in N. crassa fungi by Srb et al 1944 [15]. The authors used radiation to induce mutations in the cells, and then performed a genetic screen to isolate those with mutations relevant to the arginine synthesis pathway. This was accomplished by growing colonies of mutants in a medium which included arginine, and then in one which lacked arginine. Cells which grew in an arginine-containing medium but not without it were deemed incapable of synthesizing their own arginine, and were subsequently grown under four different conditions:

  1. In a medium lacking ornithine, citrulline, and arginine.
  2. The same medium as 1, except supplemented with ornithine only (no citrulline or arginine).
  3. The same medium as 1, except supplemented with citrulline only (no ornithine or arginine).
  4. The same medium as 1, except supplemented with arginine only (no ornithine or citrulline).

The results were as follows: 

Image c/o Biological Science 4th ed [16].

This implied that there were three types of mutants. Some had mutations preventing them from producing functional copies of the enzyme responsible for catalyzing the reaction to produce ornithine from its precursor, some for the enzyme responsible for catalyzing the reaction to produce citrulline from ornithine, and some for the enzyme responsible for catalyzing the reaction to produce arginine from citrulline. This is a simple textbook example, but the point here is that supplementing cells with the end product of a metabolic pathway negates the need for the cell to synthesize it itself through that pathway. This particular example used bread mold, but the same principle applies to bacteria.

Moreover, the Shikimic acid pathway is also metabolically expensive, so it’s not likely that the bacteria are actively using this pathway in the presence of abundant aromatic amino acids (i.e. phenylalanine, tyrosine, and tryptophan), especially when they are in competition with other microbes [17]. So, unless the person (the host) is literally starving to death, then it is far more likely gut bacteria are taking them in the easy way by just absorbing them from their environment.

If the host actually is literally starving to death or suffering from severe malnutrition, then they have far bigger and more urgent problems to worry about than their gut microbiome. Starvation and severe malnutrition themselves cause harm [18]. Consequently, parsing out and identifying harm to the host attributable to malnutrition and distinguishing it from harm to the host due to glyphosate-induced alterations to the microbiome would be problematic, especially considering that any hypothetical problems caused by the latter would be avoided by mitigating or preventing the former.

None of this is new or controversial, which is part of the reason why researchers never bothered with a full blown in vivo experiment until recently on the effects of glyphosate on the microbiome. It is also the reason why the results of the recent study should not be surprising.

Earlier Studies

Earlier studies on glyphosate’s effects on bacteria were either full of methdological problems, and/or not setup in such a way as to test the question of how it affects the microbiome in vivo, where aromatic amino acids are abundant. I’ll start with the lowest hanging fruit before dealing with more credible studies, for which the strengths and limitations are more subtle.

Samsel and Seneff

Computer scientist Stephanie Seneff is an anti-vaccine, anti-GMO, and anti-glyphosate activist who claims that GMO foods cause concussions and suggests that glyphosate in vaccines have contributed to school shootings and the Boston Bombing [19],[20]. Seriously, you can’t even make this shit up, but I digress. She and her co-author, a retired consultant by the name of Anthony Samsel, published a series of papers in a predatory pay-to-play journal (entropy) implicating glyphosate in a whole host of conditions (including celiac disease, MS, Parkinson’s, cancer, and autism), many of which involved convoluted non-sequitur arguments based on glyphosate’s alleged effects on the microbiome [21],[22]. Eric from Skeptoid has meticulously broken down the plethora of flaws and red flags in that paper, which would take way too long to reiterate here [23]. To get an idea of just how terrible that paper is, Thoughtscapism points out that it has actually been used as an example of how to spot bogus science journals: a little factoid I found far too hilarious to omit [24],[22].

Other Earlier Studies

This 1986 study showed significant growth inhibition, but only at glyphosate concentrations on the order of a millimolar or more, which is thousands of times the amounts realistically occurring in the gut from food [25]. To put this into perspective, legumes are the food crop with the highest allowed pesticide residue limit in the US (5.0 ppm) [26]. 5.0 ppm = 5.0 mg of glyph/kg of legumes, and glyphosate has a molar mass of 169.07 g/mol.

So, if we estimate that an average full stomach is roughly 1 L in volume while assuming homogeneous distribution, then we get that millimolar concentrations in the gut would involve (1 L)*(10^-3 mol of glyph/L)*(169.07 g of glyph/mol of glyph)*(10^3 mg/g) = 169.07 mg of glyphosate.

If we then assume the maximum permissible amount of glyphosate on the food crop with the highest maximum allowable glyphosate residue limit, we can calculate that millimolar concentrations in the gut by dividing the mass of glyphosate required to achieve millimolar concentrations by the mass of glyphosate per unit of mass of legumes at the maximum allowable residue limits.

When we do that, we find that it would require ingesting about 33.8 kg of legumes (or about 74.5 lbs).

i.e. (169.07 mg glyph)/(5.0 mg of glyph/kg of legumes) = 33.8 kg of legumes.

This of course assumes 100% absorption, which, as neuroscientist/geneticist/toxicologist, Alison Bernstein (aka Mommy, PhD) explains here, is actually not the case. So, the actual amount of legumes required to reach such concentrations in the gut may actually be many times higher than my sample estimate.

As Thoughtscapism points out, even at those extreme doses, the bacteria were not killed, but rather grew at a slower rate, and even that effect was partially mitigated when the researchers supplemented the bacteria with aromatic amino acids to simulate conditions likely to occur in the gut [27]. This 2010 study suffered from similar limitations [28].

Similarly, the following study showed a significant reduction in colony forming units (CFU) in vitro, but the concentrations were again on the order of a millimolar (and up to 29.5 mM), and no aromatic amino acids were supplemented to any of the test groups, which again means that it cannot be extrapolated to the gut microbiome where aromatic amino acids are abundant [29].

The Danish Study

In the new study, researchers from Denmark mapped the microbiome of Sprague Dawley rats using next generation sequencing techniques both before and after exposure both to high doses glyphosate and a commercial glyphosate formulation [10]. The researchers found that even doses 50 times that European Acceptable Daily Intake value (ADI = 0.5 mg/kg of body mass) had limited effects on microbiome composition over the course of two weeks, and that glyphosate’s effects on prototrophic bacteria growth was highly dependent on the availability of aromatic amino acids in the intestinal environment. If you are thinking that two weeks isn’t very long, you have to consider the fact that the average generational time for bacteria is roughly on the order of about 20-30 minutes (or often even less). That means that two weeks represents something on the order of (2 wks)*(7 days/wk)*(24 hrs/day)*(2-3 generations/hr) = 672–1,008 generations. Given the life expectancy of Sprague Dawley rats relative to humans, this duration is also comparable to roughly a year and a half in the life of a human [30].

What this means is that anyone continuing to promote the wrongheaded argument that glyphosate can affect health by altering the composition of the microbiome will have to hypothesize a completely new mechanism by which this is supposed to occur (preferably a biologically plausible one). This is because the reasoning behind this argument is based on the premise that glyphosate-induced inhibition of the shikimic acid pathway in gut microorganisms should prevent them from growing due to their (wrongly) assumed dependence on it for the synthesis of aromatic amino acids. This hypothesis predicts that hundreds of generations of bacteria should not be permitted to grow normally if this effect is occurring to any meaningful degree. The evidence falsifies this prediction.


The claim that glyphosate harms human health via disruption of the microbiome was never a biologically plausible one, because it only makes sense when the system is not being viewed as a whole. Ironically, glyphosate and GE food opponents like to say that they take a holistic approach, but this is not a holistic argument, because it ignores the environment in which the microbiome exists.

We know that organisms don’t bother synthesizing compounds they can already get from their environment. Knocking out one step of a biochemical pathway and growing microorganisms on different media with various substrates is a tried and true classical method for identifying which substrates are involved in a given pathway and/or the enzymes which catalyze their reactions. We also know that the human gut contains abundant aromatic amino acids alleviating the need for resident microorganisms to synthesize them. Running out of them is not a concern because they are replenished multiple times per day. The exception to this would be cases of starvation or malnutrition, in which case malnutrition would be the problem to address: not glyphosate. Despite this, in vivo research has been done, and reaffirms exactly what theoretical predictions would imply. Gut microorganisms grew and replicated for hundreds of generations, thus contradicting the predictions of the hypothesis under discussion.

In order to continue to argue that glyphosate had some other negative effect on the microbiome which would be undetectable within the first several hundred or more generations, a contrarian would have to either postulate a different mechanism by which this could be rendered into a testable scientific hypothesis, or appeal to vague and unspecified unknowns.

In the former case, this would constitute an abandonment of the original argument in place of a new hypothesis leading to predictions distinct from those of the hypothesis under discussion. Essentially, this would mean conceding (either explicitly or implicitly) that the original claim was false (or at least not supported), and then moving the goalpost to a new claim based on a different mechanism.

In the latter case, such vague and half-baked speculation could be applied just as easily to virtually anything. It makes no specific postulates and thus makes no testable predictions, and is therefore unscientific. It is what we sometimes refer to as “not even wrong” [31].

– Cred Hulk

For more on glyphosate and common myths about it, Thoughtscapism has put together the most comprehensive piece I’ve ever seen on the subject for a general audience [32].


[1] Glyphosate | History of glyphosate. (2017). Retrieved 10 December 2017, from

[2] (2017). Retrieved 10 December 2017, from

[3] Hulk, C. (2015). Glyphosate toxicity: Looking past the hyperbole, and sorting through the facts. By Credible HulkThe Credible Hulk. Retrieved 10 December 2017, from

[4] Scientific evidence that Roundup is dangerous has been mounting.. (2017). Greenpeace International. Retrieved 10 December 2017, from

[5] Millions march against GM crops. (2013). the Guardian. Retrieved 10 December 2017, from

[6] Glyphosate | Glyphosate: mechanism of action. (2017). Retrieved 10 December 2017, from

[7] Starcevic, A., Akthar, S., Dunlap, W. C., Shick, J. M., Hranueli, D., Cullum, J., & Long, P. F. (2008). Enzymes of the shikimic acid pathway encoded in the genome of a basal metazoan, Nematostella vectensis, have microbial origins. Proceedings of the National Academy of Sciences105(7), 2533-2537.

[8] Sammons, R. D., Gruys, K. J., Anderson, K. S., Johnson, K. A., & Sikorski, J. A. (1995). Reevaluating glyphosate as a transition-state inhibitor of EPSP synthase: Identification of an EPSP synthase. cntdot. EPSP. cntdot. glyphosate ternary complex. Biochemistry34(19), 6433-6440.

[9] Alibhai, M. F., & Stallings, W. C. (2001). Closing down on glyphosate inhibition—with a new structure for drug discovery. Proceedings of the National Academy of Sciences98(6), 2944-2946.

[10] Nielsen, L. N., Roager, H. M., Frandsen, H. L., Gosewinkel, U., Bester, K., Licht, T. R., … & Bahl, M. I. (2018). Glyphosate has limited short-term effects on commensal bacterial community composition in the gut environment due to sufficient aromatic amino acid levels. Environmental Pollution233, 364-376.

[11] Hudson, A. O., Harkness, T. C., & Savka, M. A. (2016). Functional Complementation Analysis (FCA): A Laboratory Exercise Designed and Implemented to Supplement the Teaching of Biochemical Pathways. JoVE (Journal of Visualized Experiments), (112), e53850-e53850.

[12] Sohaskey, C. D., & Wayne, L. G. (2003). Role of narK2X and narGHJI in hypoxic upregulation of nitrate reduction by Mycobacterium tuberculosis. Journal of bacteriology185(24), 7247-7256.

[13] Smits, T. H., Balada, S. B., Witholt, B., & van Beilen, J. B. (2002). Functional analysis of alkane hydroxylases from gram-negative and gram-positive bacteria. Journal of bacteriology184(6), 1733-1742.

[14] Salcedo, E., Cortese, J. F., Plowe, C. V., Sims, P. F., & Hyde, J. E. (2001). A bifunctional dihydrofolate synthetase–folylpolyglutamate synthetase in Plasmodium falciparum identified by functional complementation in yeast and bacteria. Molecular and biochemical parasitology112(2), 239-252.

[15] Srb, A., & Horowitz, N. H. (1944). The ornithine cycle in Neurospora and its genetic control. Journal of Biological Chemistry154(1), 129-139.

[16] Freeman, S. (2017). Biological Science (6th ed.). Edinburgh Gate Harlow Essex CM20 2JE England. Pearson Education.

[17] Hibbing, M. E., Fuqua, C., Parsek, M. R., & Peterson, S. B. (2010). Bacterial competition: surviving and thriving in the microbial jungle. Nature Reviews Microbiology8(1), 15-25.

[18] Correia, M. I. T., & Waitzberg, D. L. (2003). The impact of malnutrition on morbidity, mortality, length of hospital stay and costs evaluated through a multivariate model analysis. Clinical nutrition22(3), 235-239.

[19] Seneff Claims GMOs Cause Concussions. (2015). Science-Based Medicine. Retrieved 10 December 2017, from

[20] Who is Stephanie Seneff?. (2017). VAXOPEDIA. Retrieved 10 December 2017, from

[21] Anthony Samsel (n.d.) LinkedIn [Profile page]. Retrieved Dec 10. 2017, from

[22] A guide to detecting bogus scientific journals. (2015). Sci-Phy. Retrieved 10 December 2017, from

[23] Roundup and Gut Bacteria. (2013). Skeptoid. Retrieved 10 December 2017, from

[24] →, V. (2016). 2.-3. Glyphosate and Health Effects A-ZThoughtscapism. Retrieved 10 December 2017, from

[25] Fischer, R. S., Berry, A. L. A. N., Gaines, C. G., & Jensen, R. A. (1986). Comparative action of glyphosate as a trigger of energy drain in eubacteria. Journal of bacteriology168(3), 1147-1154.

[26] (2017). Retrieved 10 December 2017, from

[27] →, V. (2016). 4. Does Glyphosate Harm Gut Bacteria?Thoughtscapism. Retrieved 10 December 2017, from

[28] Ahemad, M., & Khan, M. S. (2011). Toxicological effects of selective herbicides on plant growth promoting activities of phosphate solubilizing Klebsiella sp. strain PS19. Current microbiology62(2), 532-538.

[29] Shehata, A. A., Schrödl, W., Aldin, A. A., Hafez, H. M., & Krüger, M. (2013). The effect of glyphosate on potential pathogens and beneficial members of poultry microbiota in vitro. Current microbiology66(4), 350-358.

[30] Andreollo, N. A., Santos, E. F. D., Araújo, M. R., & Lopes, L. R. (2012). Rat’s age versus human’s age: what is the relationship?. ABCD. Arquivos Brasileiros de Cirurgia Digestiva (São Paulo)25(1), 49-51.

[31] Burkeman, O. (2005). Briefing: Not even wrongthe Guardian. Retrieved 10 December 2017, from

[32] 17 Questions About Glyphosate. (2016). Thoughtscapism. Retrieved 10 December 2017, from

[33] Wang, Y., & Kasper, L. H. (2014). The role of microbiome in central nervous system disorders. Brain, behavior, and immunity38, 1-12.

[34] Germ theory denialism and the magical mystical microbiome – RESPECTFUL INSOLENCE. (2015). RESPECTFUL INSOLENCE. Retrieved 10 December 2017, from

[35] Forbes Welcome. (2017). Retrieved 10 December 2017, from

[36] Gut Check. Probiotics and Metabiome.. (2015). Science-Based Medicine. Retrieved 10 December 2017, from

[37] Zucko, J., Dunlap, W. C., Shick, J. M., Cullum, J., Cercelet, F., Amin, B., … & Long, P. F. (2010). Global genome analysis of the shikimic acid pathway reveals greater gene loss in host-associated than in free-living bacteria. BMC genomics11(1), 628.


A chemical is a chemical is a chemical

A chemical’s properties are not affected by whether it was made by God, Mother Nature, evolution, or a chemist in a lab coat.

There’s a certain irony to drinking a soy latte from a BPA-free mug, but that’s not something the chemophobic fear squad seems to understand. This fear-mongering crowd would have you believe that synthetic endocrine-disrupting chemicals (EDCs) are going to be the downfall of civilization as we know it (I admit, that’s hyperbolic, but so are many of their exaggerated claims). However, in a truly perfect example of inconsistent application of the precautionary principle and selective consideration of data and data gaps, they completely overlook the potential risks of phytoestrogens, naturally occurring endocrine-disrupting chemicals found in soy and other common foods.

Phytoestrogens are of particular concern because of the increased consumption of phytoestrogen-containing food. Sales of soy-based food products (tofu, tempeh, edamame, soy “milk”, soy “yogurt”, soy-based baby formula) have grown substantially since the late 1990s. Other sources of phytoestrogens include grapes and red wine, citrus fruits and juices, parsley, celery, pepper, kale, broccoli, onions, tomatoes, lettuce, apples, chocolate, green tea, beans, apricots, cherries, berries, spinach, and flax seed.

Data suggest that the exposure to EDCs from natural dietary sources may be higher than exposure to synthetic EDCs although this is a difficult comparison to make conclusively. Epidemiological studies of the health effects of both show mixed, complicated (i.e. confounded by associated factors), and mostly small results. Despite this nuance, we tend to view natural EDCs favorably, and synthetic EDCs unfavorably, despite having similar chemical properties. Source, not evidence, is the foundation for the difference in attitude toward natural and synthetic EDCs.

Now, this is not meant to make you afraid of eating soy, kale, berries and all these other foods, or of using your shampoo. I point this out to draw attention to the fact that that we tolerate a lot of risk in our lives and that exposure and risk are unavoidable parts of life. A simplistic dichotomy between the hazards of synthetic exposures and safety of natural exposures is often not based on evidence. There is truly no such thing as a risk-free, exposure-free life, even with so-called “all-natural” products. However, there are some simple and easy steps to reduce exposures if you are concerned, like avoiding soy-based products and use less plastic in food preparation and storage during sensitive developmental periods.

Reducing and mitigating risks where it is feasible to do so is reasonable, but it is often not a simple undertaking. For example, a recommendation to completely avoid phytoestrogen containing food would be misguided as evidence consistently shows that, despite concerns about exposure to phytoestrogens at certain times in life, a diet rich in a wide variety of fruits and vegetables produces a net benefit for health.

The similarities between natural and synthetic chemicals are often overlooked as we fall victim to our assumptions that natural is good and synthetic is bad. A chemical’s properties are not affected by how it was produced. Synthetic chemicals that have a biological effect are able to do so because they are able to mimic or interact with systems that exist in nature. Chemistry, not source, determines how a chemical acts in a biological system.

From a 2010 review, “The pros and cons of phytoestrogens”:

Phytoestrogens are intriguing because, although they behave similarly to numerous synthetic compounds in laboratory models of endocrine disruption, society embraces these compounds at the same time it rejects, often with vigor, use of synthetic endocrine disruptors in household products. Thus, phytoestrogens both expand our view of environmental endocrine disruptors and propound that the source of the compound in question can influence the direction and interpretation of research and available data. While the potentially beneficial effects of phytoestrogen consumption have been eagerly pursued, and frequently overstated, the potentially adverse effects of these compounds are likely underappreciated. The opposite situation exists for synthetic endocrine disruptors, most of which have lower binding affinities for classical ERs than any of the phytoestrogens but can sometimes produce similar biological effects. (emphasis mine)

In interpreting information about health effects of chemicals, wherever they come from, we must be aware of our biases in favor of the natural and against the synthetic to ensure that we analyze data objectively. Only when we consider evidence without bias can we arrive at sound recommendations and regulations.EDCs



About those harsher herbicides that glyphosate helped replace:

One of the common criticisms of commercially available Genetically Engineered (GE) seeds is the idea that they have led to an increase in pesticide use. In actuality, it turns out that they’ve corresponded to a decrease in total pesticide use, but this is attributable primarily to insect resistant GE crops, and critics argue that herbicide resistant crops have led to an increase in herbicide usage. It is true that the rise in popularity of glyphosate-resistant (GR) crops in particular has coincided with an increase in the use of glyphosate, which had already been in use to some degree for a couple of decades before the implementation of glyphosate-resistant crops. However, what critics invariably fail to mention is that its rise in popularity also coincided with the phasing out of other herbicides, most of which were significantly more toxic than glyphosate (about which I’ve written in detail here).

The purpose of this article is not to claim that glyphosate and GR crops are the be all end all of weed control (they’re not), nor is it to claim that they were causally responsible for any and every desirable change we see in herbicide usages patterns. Rather, the purpose of this is to show that when opponents of GE technology and of glyphosate claim that GR crops are bad on the grounds that they increased glyphosate use, they are leaving out critical information that would be highly inconvenient for their narrative.

It’s important to note that the data upon which these usage timeline graphs are based is very USA-centric. Perhaps a timeline and analysis of herbicide usage patterns in other places would be a good topic for another article, but the US is not a bad place to start because we do cultivate a lot of glyphosate-resistant crops here, as well as a lot of GE crops in general.

What were some of these herbicides?

Alachlor was one of them. The EPA states the following about alachlor:

“The greatest use of alachlor is  as a herbicide for control of annual grasses and broadleaf weeds in crops, primarily on corn, sorghum and soybeans.”

Alachlor use

Alachlor: [Source]

According to the EPA’s Water division:

“Some people who drink water containing alachlor well in excess of the maximum contaminant level (MCL) for many years could have problems with their eyes, liver, kidneys, or spleen, or experience anemia, and may have increased risk of getting cancer.”

The EPA and OSHA list alachlor as a Class L1 Carcinogen, which means they consider it likely to be carcinogenic at high doses but not at low doses. With an LD50 of between 930 mg/kg and 1,350 mg/kg in rats, and between 1,910 and 2,310 mg/kg in mice, its acute toxicity is not generally considered to be a big concern (although you may notice that it is still noticeably more acutely toxic than glyphosate which has an LD50 of 5,600 mg/kg). However, its potential for chronic toxicity remains a concern, particularly for the liver, spleen and kidneys (according to its PMEP profile) , and its NOAEL varied depending on the duration of the study in question.

Okay then. What else? How about Cyanazine?

Cyanazine use dropped to zero

Cyanazine use dropped to zero shortly after the rise of GR crops.

Cyanazine has an LD50 of between 182 and 332 mg/kg in rats and 380 mg/kg in mice (far more acutely toxic than glyphosate or alachlor), but its long term effects and NOAEL varied from anywhere around 0.198 to 3.3 mg/kg depending on which study you look at (as you may read more about in this WHO report). Although cyanazine is not known to be carcinogenic for certain, it has been observed to affect the central nervous system upon over-exposure and to increase liver weight while decreasing body weight gain.

Cyanazine was eventually put under special review due to concerns over its possible cancer-causing potential. DuPont voluntarily discontinued it in 1999, and its sale in the US was officially prohibited by 2002.

So, needless to say, cyanazine use went way down rather abruptly. What else?  According to PMEP, Fluazifop “is a selective phenoxy herbicide used for postemergence control of annual and perennial grass weeds. It is used on soybeans and other broad-leaved crops such as carrots, spinach, potatoes, and ornamentals.”

Fluazifop usage

Fluazifop usage

Fluazifop hasn’t gone away completely, but its use did decline quite significantly, possibly thanks in part to the introduction of glyphosate resistant soybeans. How toxic is it though? its LD50 was 3,680 for male rats and 2,451 for female rats, which is only a little bit more acutely toxic than glyphosate, but PMEP notes the following:

“A single dose of the formulated compound (Fusilade 2000) can cause severe stomach and intestine disturbance. Ingestion of large quantities may cause problems in the central nervous system such as drowsiness, dizziness, loss of coordination and fatigue. Breathing small amounts of the product may cause vomiting and severe lung congestion. This may ultimately lead to labored breathing, coma and death.”

So, yeah. There’s that. The good news is that there was no evidence of chronic toxicity in rats under 10 mg/kg per day in 90 day trials.

The next one up is metolachlor.

Technical grade Metolachlor has an LD50 of between 1,200-2780 mg/kg in rats. That’s between twice and 4.67 times the acute toxicity of glyphosate. Additionally, with an NOAEL of roughly 90 mg/kg/day, metolachlor can exhibit chronic toxic effects at doses MUCH smaller than the levels at which it becomes acutely toxic. Symptoms of  human intoxication from metolachlor include abdominal cramps, anemia, shortness of breath, dark urine, convulsions, diarrhea, jaundice, weakness, nausea, sweating, and dizziness.

Okay. Great, but what about Atrazine? In 1996 Atrazine was the #1 herbicide for corn. 2 and 3 were cyanazine and alachalor which, as we just saw, have effectively been zeroed out.

Well, apparently the rise of GR crops has had little to no effect on atrazine usage in the US. This might come as both a surprise and a disappointment to some because atrazine is known to degrade very slowly in soil (often lasting for months) and  has been known to inadvertently end up in drinking water, a fact which contributed to it being banned in the EU. It’s also a suspected endocrine disruptor and is more acutely toxic than glyphosate (with an LD50 of 672 to 3,000 mg/kg in rats). The EPA also classified it as a possible carcinogen, and multiple undesirable biochemical and morphological changes in various organs have been observed in high dose studies of its chronic toxicity. That’s probably not what most of my readers wanted to hear. However, part of being a responsible skeptic is understanding the importance of not cherry picking data. Additionally, we may be able to learn something by asking why this is the case. While at first glance this result is not so exciting, bear in mind that resistant weeds have increased quite a bit without increasing use AND corn production is greatly increased (by about 54%) since 1996, so use per bushel is down (as is use per capita because the population is up in the US by about 50 million people since then). Alright then. That’s not as spectacular as those previous examples, but at least it wasn’t a total bust.

What else? How about Metribuzin?

In 1992, over 2.5 million lbs of metribuzin was used just on soybeans alone. After that, its usage on fruits and vegetables didn’t change too drastically, but we can see that its use on soybeans and its overall use dropped dramatically. It eventually started climbing back up, and there a number of possible reasons for that, but it did initially go down, particularly in soybeans (for which a glyphosate-resistant variety was introduced by Monsanto in 1996). Metribuzin’s LD50 is 1,090 to 2,300 mg/kg in rats, which is about 2.5 to 5 times as toxic as glyphosate. None of the studies looking at chronic toxicity revealed any negative effects at any of the dosages tested.

Another one that was popular in the mid 90s in the US was Nicosulfuron.

Nicosulfuron usage

Nicosulfuron usage

The acute toxicity of Nicosulfuron is not much worse than glyphosate, with an estimated LD50 of in excess of 5,000 mg/kg of body mass. As for chronic toxicity, its NOAEL was found to be about 125 mg/kg/day, and its LOAEL (Lowest Observed Adverse Effect Limit) was found to be 500 mg/kg/day according to the EPA.

If you’ve made it this far, you may be wondering how this data was obtained. The website was created by USGS National Water Quality Assessment (NAWQA) Program.

“The pesticide-use maps provided on this web site show the geographic distribution of estimated use on agricultural land in the conterminous United States for numerous pesticides (active ingredients). Maps were created by allocating county-level use estimates to agricultural land within each county. A graph accompanies each map, which shows annual national use by major crop for the mapped pesticide for each year.

Methods for generating county-level pesticide use estimates are described in Estimation of Annual Agricultural Pesticide Use for Counties of the Conterminous United States, 1992–2009 (Thelin and Stone, 2013) and Estimated Annual Agricultural Pesticide Use for Counties of the Conterminous United States, 2008-12 (Baker and Stone, 2015).  Two different methods, EPest-low and EPest-high, were used to estimate a range of use, with the exception of estimates for California, which were taken from annual Department of Pesticide Regulation Pesticide Use Reports (Baker and Stone, 2015).”

A PDF copy of the entire Thelin and Stone report can be found here, but their quick summary of the data sources is as follows:

“Data sources used to develop EPest pesticide-by-crop use rates and annual pesticide-use estimates by county included the following: (1) proprietary pesticide-by-crop use estimates reported for CRDs; (2) USDA county harvestedcrop acreage reported in the 1992, 1997, 2002, and 2007 Census of Agriculture (, and NASS annual harvested-crop acreage data collected from crop surveys for non-census years (http://quickstats.nass.usda. gov/); (3) boundaries for CRDs and counties; (4) regional boundaries derived from USDA Farm Resource Regions; and (5) pesticide-use information from California DPR-PUR. Each of these sources is described in following sections.”

The USGS also includes this statement on the strengths and limitations of the data:

“Pesticide use estimates from this study are suitable for making national, regional, and watershed assessments of annual pesticide use, however the reliability of estimates generally decreases with scale.  For example, detailed interpretation of use intensity distribution within a county is not an appropriate use.  Although county-level estimates were used to create the maps and are provided in the dataset, it is important to understand that surveyed pesticide-by-crop use was not available for all CRDs and, therefore, extrapolation methods were used to estimate pesticide use for some counties. Surveyed pesticide-by-crop use may not reflect all agricultural use on all crops grown. In addition, state-based restrictions on pesticide use were not incorporated into EPest-high or EPest-low estimates. EPest-low estimates are more likely to reflect these restrictions than EPest-high estimates. With these caveats in mind, including other details discussed in Thelin and Stone (2013) and Baker and Stone (2015), the maps, graphs, and associated county-level use data are critical data for water-quality models and provide a comprehensive graphical overview of the geographic distribution and trends in agricultural use in the conterminous United States.”

Many people never even hear about the herbicides that were phased out in favor of glyphosate simply because they aren’t pertinent to the anti-agricultural biotech narrative, and because their popularity had waned by the time it had become trendy to demonize GMOs and everything remotely associated with them.

I said this before, and I’ll say it again:

“Opponents of glyphosate often seem to hold this unfounded notion that, if they can manage to get glyphosate banned or simply willingly abandoned, then it would mean an improvement in both food and environmental safety, but the truth is it would likely be the exact opposite of that. Weeds are a legitimate problem in farming that has to be dealt with one way or another. In its absence, it would have to be replaced with something else, and it would likely be something more caustic: not less.”


– Credible Hulk.

Check out my Facebook page. 

I also help run Stop the Anti-Science Movement.