Omitting the Science: A Critique of “Rigging the Science of GMO Ecotoxicity”
Recently, I have been seeing an article titled “Rigging the Science of GMO Ecotoxicity” making the rounds. This article is written by Dr. Latham on his website “Independent Science News”, part of the “Bioscience Resource Project”, cofounded by Latham. The focus of the article is on Bt crops, one of the most common types of transgenic (commonly called “GMO”) traits. There are two major claims of the article. The first claim is that Bt is toxic to non-target insects (ones that don’t eat the crop). The second claim is that companies are hiding this fact by using a trick to manipulate safety studies. As worrying as those claims sound, they simply aren’t true. The article misleads readers by omitting a lot of facts, hence my clumsy title “Omitting the Science”. The issues with Latham’s article are many and can get technical, but I can briefly summarize these in four points, which are described in detail after:
1) Cherry-picking studies: the article completely ignores 20+ years of research contradictory to their claims.
Specific example: The article mentions that Bt corn pollen was lethal when force fed to Monarch larvae, but does not mention that field studies have shown little risk to Monarchs[3, 4].
2) Cherry-picking results: in the studies that are cited, the article ignores results are contradictory to their claims.
Specific example: The article cites Porcar et al. (2010) to support the claim that antibiotics can modify the effect of Bt. but never once mention that in Porcar et al.’s experiments they also found that Green Lacewings were unaffected by Bt, even in the absence of antibiotics.
3) Overstating results: the article overstates results to support their claims.
Specific example: The article cites Hilbeck et al. (2018) as showing that antibiotics modifies the effects of Bt, but fail to mention that this effect was small. In fact, Bt still had a lethal effect in all experiments and in one experiment antibiotics actually seemed to increase lethality.
4) Claims a nefarious conspiracy to hide ecotoxicity.
Specific example: They imply that the use of antibiotics in feeding studies is a nefarious conspiracy by industry, but present no evidence of this.
These are major issues of omission. Omissions that mislead people about the safety of Bt crops and the rigors of safety assessments. For those interested in the specifics of these issues, please read on.
What are Bt crops?
Bt plants were developed to be resistant to insect pests that eat the crop. They do this by producing “Cry” proteins, a type of protein that is toxic to specific families of insects, but otherwise does not affect other species, including mammals. These proteins come from certain soil bacteria called Bacillus thuringiensis and are also found in the guts of certain caterpillars. Discovered in 1901, these bacteria were first used as a natural pesticide in Agriculture in the 1920s, ~70 years before the first commercial transgenic plants. It is still used in this way to this day, including as a natural pesticide on organic farms. The safety of Bt has been tested for decades, including direct human tests with humans in 1959. In the 80s and 90s, the genes for Cry proteins, which gives Bacillus thuringiensis this unique ability were expressed in crops like cotton and corn. This gave these crops resistance to major pests like Cotton Bollworm or European Corn Borer, that cause significant damage and losses to farmers. This also allowed farmers to stop spraying many other pesticides previously used to control these pests, pesticides that often affected non-target insects.
Issue 1: Cherry-picking studies
When I first read Latham’s article, the very first red-flag that something was not quite right was this claim:
“Organisms affected by Cry toxins include monarch butterflies, swallowtail butterflies, lacewings, caddisflies, bees, water fleas, and mammals”
I knew right away this was wrong, as I still remember the debate about Bt corn and Monarch butterflies that occurred in the early 2000s and knew there was much that Latham left out. The controversy arose 20 years ago, when a paper was published by Losey et al. (1999) claiming that Bt corn pollen was toxic to Monarch caterpillars. However, Monarch caterpillars don’t eat corn pollen, they primarily eat milkweed. So to show this, Losey et al. essentially force-fed some Monarchs in the laboratory to show that it was indeed toxic. Part of the argument was that pollen from corn fields could end up on milkweed leaves and thus be accidentally ingested by the caterpillars.
However, not long after this, a number of studies set out to put these claims to the test using actual field experiments. These papers, published in 2001, showed that in actual field conditions, there was little to no risk of such a scenario[3, 4]. All of this research has been known and public for some time. The field studies that showed no risk to Monarchs has been public knowledge for 18 years, plenty of time for Latham to have become familiar with this important research. Despite this, Latham’s article only cites the single Losey et al. (1999) study, conveniently ignoring years of research that does not support the narrative he is trying to push. Ironically, not long after Latham’s article went online, another study was published, showing that declines in Monarch populations and the milkweed they feed on, predate the introduction of Bt corn by decades.
It’s not just the claims made about Monarchs either. Similar work in other organisms, like swallowtail butterflies, lacewings, etc. have also shown little risk to non-target insects. Even more astounding are the claims that Cry proteins can impact human and animal health. Again to cite Latham:
“They suspect that Cry toxins may affect many non-target species, even including mammals and humans (e.g. Dolezel et al., 2011; Latham et al., 2017; Zdziarski, et al., 2018).”
This is again quite a selective reading list and the second of these papers. Latham’s own paper in that mix, doesn’t even present data supporting such a claim of harm, but is ridden with many of the same issues I state here (perhaps I will address these in a future post). Recall that earlier I mentioned how Bt and Cry proteins have been studied for decades, even including direct human safety tests. While some published studies have reported negative effects in animal feeding studies, these have their own issues and fail to replicate. For example, recently has been a series of independent and rigorous rat feeding studies conducted by the EU[15–17], providing no evidence of risk to animals, let alone humans. Indeed, in 20+ years of transgenic plants, the overwhelming evidence shows no harm to animals or humans[7, 9, 18].
These are major oversights, deliberate or not. The studies referenced are cherry-picked to support an agenda, not enlighten readers or engage with the science. I understand that Latham’s article is not intended for scholarly peer-review, where such standards are expected. Instead, it is intended for a more general audience, which makes these omissions all the worse. At least in a peer-review article, Scientists will hopefully have the background to know of decades of research omitted, but for a general audience, it is simply misleading. It falsely gives readers the view that the claims made by Latham are settled, settled to support his claims, when in fact much of the literature says the opposite.
Issue 2: Cherry-picking results
I realize that this next issue “Cherry-picking results” sounds and is indeed much like the first issue “Cherry-picking studies”, but I think it deserves to stand on its own. First, while one could give Latham the benefit of the doubt regarding his familiarity with all the studies he fails to cite, it’s quite another matter to ignore results from the very papers you are citing and even quoting. I don’t think it is much to ask, that if people are going to cite a paper, they actually read it, try to understand it, and accurately represent it.
After reading Latham’s article, I looked up each and every one of his references and read them. In particular, I was interested in the discussion of the use of antibiotics in insect feeding studies. One study in particular was interesting, that was Porcar et al (2010). Latham directly quotes a small section of this paper, showing that he read it.
“Antibiotics were deliberately excluded from the diet composition since bacteria occurring in the insect midgut naturally might be critical for sensitivity (Broderick et al., 2006).” (Porcar et al. 2010)
However, having quoted Porcar et al. (2010) on this issue, Latham makes no further mention of this study. What was this paper about? It was an insect feeding study testing the effects of Cry proteins on Green Lacewings, a non-target insect. The photo at the beginning of Latham’s article is in fact of a Green Lacewing. These are beneficial insects, which prey on many plant pests, and so it is important to minimize potential negative effects on them.
While Latham criticizes several studies that also looked at Green Lacewings for using antibiotics in their feeding, saying that this will “mask” the lethal effects of Cry proteins, he never once mentions that Porcar et al. (2010), also found no effect of Cry proteins on Green Lacewings. This is quite astounding, as it independently verifies the results of the studies that did include antibiotics, so why doesn’t Latham mention this? Surely it would be useful to readers to know that Cry proteins are safe for this important species, regardless of whether the study used antibiotics or not? Instead readers are left to assume that Cry proteins are harmful and that antibiotics were included in the study to mask this effect. Such omissions, even of the very research Latham cites and shows he has read, mislead readers.
Issue 3: Overstating results
Now we come to Hilbeck et al. (2018), the first paper cited by Latham and the catalyst for his article. The first author of the paper, Dr. Hilbeck is a collaborator and co-author of Latham . This study conducts a number of feeding studies, with and without antibiotics, using two different insect species, European Corn Borer and Cotton Leafworm. Unlike Green Lacewing or Monarchs, these are crop pests and the intended targets of Bt crops. The Cry proteins are supposed to protect crops from these insects. In the feeding studies, the authors do observe a difference in lethality whether insects are fed antibiotics or not. I do not dispute this, indeed based on previous research, I have no doubt that antibiotics can modify the effects of Cry proteins on some insects. I would like to emphasize the “some insects”, because as we saw with Porcar et al. (2010), antibiotics seem not to be a factor for species like Green Lacewing.
In science, if a study is correctly designed, we can often detect small differences, even if their overall effect is small. This does not mean they are non-existent, but it does mean that we shouldn’t overstate their importance either. If I lost my wallet, the impact is quite a bit different if my wallet contained $1 or $1000. While Hilbeck et al. (2018) do see an effect of antibiotics on Cry effectiveness in two target species, if one looks at the actual data that effect is actually quite small. I will not show the actual figures, due to the risk of copyright issues, but the paper is open access for anyone who wants to check for themselves (https://www.mdpi.com/2072-6651/10/12/489/htm). Even at the highest antibiotic dose, it’s very obvious that Cry proteins are still lethal to the two target species and distinguishable from the control diet with no Cry proteins. Interestingly, in one experiment, the diets with antibiotics seemed more effective than the diet without. Such a small effect of antibiotics in the diet are not a hindrance to testing the effects of Cry proteins and something that any well designed study would be able to control for.
Despite this, Latham, uses Hilbeck et al. (2018) as a launch pad to claim that antibiotics “mask” the effects of Cry proteins and so “rig” the science of eco-toxicological studies. But that is to overstate the case. If you can still detect lethal effects of Cry proteins with or without antibiotics in the diet, you are not “masking” anything. If as with Porcar et al. (2010), you see no difference in results with or without antibiotics, then again, you are not “masking” anything. Taking relatively small effects and arguing that the entire field, decades of research, are flawed, is not only overstating the claim, it’s just plain misleading.
Issue 4: Conspiracy Theories
Issues 1–3 dealt with issues of scholarship, namely the omission of studies, results, and details that put the whole issue Bt crops and their ecotoxicology into perspective. They demonstrate how by omitting these facts, readers are misled about the science. All of this omission has been to build support for the claim Cry proteins are a major ecological risk and that this fact is being hidden…“masked”…by companies through the use of antibiotics in feeding studies.
This is conspiratorial thinking, something which we should reject from the outset. No evidence of such a conspiracy is presented. To push this narrative, Latham targets three papers for criticism[19–21]; criticizing both the amount of antibiotics used and the justification for their use. What was that justification? If you dig through the literature, it appears that these authors used a diet based an older academic paper on methods of rearing insects in captivity. Antibiotics in such diets are used for a couple of important reasons: spoilage and pathogens. Of these two reasons, Latham only mentions “spoilage”, omitting the importance of pathogens.
In any feeding study, it is critically important that there be high consistency in the diets you feed. Batch to batch variation can introduce uncontrolled effects, making the study less effective. Just as the food in your fridge can go bad over time, so can the food used to feed insects. Preventing spoilage, caused by say bacteria, of the food is important. For this, antibiotics are often introduced. But this is not the only reason. The environment is rife with pathogens. This goes for insects as well as people, for example the pathogenic bacteria from the genus Serratia. If these were to infect a food source or a study, not only would the entire study have to be scrapped, but the lab itself might have to stop research until the issue was dealt with. Such studies cannot use entirely sterile conditions (which introduces its own issues). An added precaution then, is to also add antibiotics to the diet. Why Latham does not mention this second issue, I don’t know, perhaps he should have consulted an entomologist with experience in these matters.
So it turns out, that there is often good practical scientific basis for the inclusion of antibiotics in insect diets. This is far from some grand conspiracy to mask the effects of Cry proteins, but common practice in many labs, both in industry and in academia. Ironically, the three papers which Latham openly criticize are not industry studies, but were conducted by university researchers, in China, funded by the Chinese government[19–21]. While the issue of antibiotics in feeding studies should be considered, it is irresponsible to inject conspiratorial thinking into such a discussion. This is an issue to be addressed through science, through careful experimental design, factors that are standard practice and which are being accounted for, as the scientific literature attests to.
I think the four issues I highlight above are serious omissions. Omissions which mislead readers. Such omissions would be bad enough in a scientific paper, but scientists often have the benefit of having access to studies and colleagues who can correct these issues that are not readily available to the general public. When the target is a general audience without easy access to these same resources, then it is all the more important that the facts be presented openly and completely.
Lest people get the wrong impression, I agree that antibiotics could indeed have an effect on the effects of Cry proteins in feeding studies and think the scientific literature has shown this. However, it seems clear, that this is not a universal effect, but can and does vary by species and potentially other unknown factors. Even then, the effects observed seem mixed, certainly not enough to justify a claim that they are “masking” ecotoxicity. It is apparent that these effects have been known in the literature for some time and that there is also often good reason to use antibiotics in such studies. These are factors that can be controlled. Ultimately, such results should be interpreted in the light of the totality of research . Risk is not determined from a single controlled laboratory experiment. Risk is a combination of potential harm and actual exposure. Risk can only be determined by accounting the totality of data, from both laboratory and field experiments[12, 23]. So while the issues raised by Latham should not be dismissed, the reality is they aren’t and we have decades of data to support the safe use of Bt crops. Omitting the science is inexcusable.
I am an academic plant biologist (just search my name if you want more detail). I have received no compensation for this article. As of the date of my writing this (February 10, 2019), I have received no funding or compensation from industry for my research, nor is any anticipated. I would like to acknowledge and thank Joe Ballenger, an entomologist with extensive expertise in insect feeding studies, for his feedback. Joe did not write any part of this article, but did inform me about issues such as the pathogenic bacteria Serratia, the laboratory practices in feeding studies, and expression levels of Cry proteins in maize pollen. Joe Ballenger has previous worked in contract with Bayer, but is no longer affiliated with them.
1. Latham, J.R. Rigging the Science of GMO Ecotoxicity. 2019 [cited 2019 February 10, 2019]; Available from: https://www.independentsciencenews.org/environment/new-evidence-of-gmo-bt-crop-safety-manipulation/.
2. Losey, J.E., L.S. Rayor, and M.E. Carter, Transgenic pollen harms monarch larvae. Nature, 1999. 399: p. 214.
3. Stanley-Horn, D.E., et al., Assessing the impact of Cry1Ab-expressing corn pollen on monarch butterfly larvae in field studies. Proceedings of the National Academy of Sciences, 2001. 98(21): p. 11931–11936.
4. Sears, M.K., et al., Impact of <em>Bt</em> corn pollen on monarch butterfly populations: A risk assessment. Proceedings of the National Academy of Sciences, 2001. 98(21): p. 11937–11942.
5. Porcar, M., et al., Effects of Bacillus thuringiensis Cry1Ab and Cry3Aa endotoxins on predatory Coleoptera tested through artificial diet-incorporation bioassays. Bulletin of Entomological Research, 2010. 100(3): p. 297–302.
6. Hilbeck, A., et al., Impact of Antibiotics on Efficacy of Cry Toxins Produced in Two Different Genetically Modified Bt Maize Varieties in Two Lepidopteran Herbivore Species, Ostrinia nubilalis and Spodoptera littoralis. Toxins, 2018. 10(12).
7. Manzo, A., F. Veronesi, and D. Rosellini, An overview of the last 10 years of genetically engineered crop safety research AU — Nicolia, Alessandro. Critical Reviews in Biotechnology, 2014. 34(1): p. 77–88.
8. Fisher, R. and L. Rosner, Insecticide Safety, Toxicology of the Microbial Insecticide, Thuricide. Journal of Agricultural and Food Chemistry, 1959. 7(10): p. 686–688.
9. Pellegrino, E., et al., Impact of genetically engineered maize on agronomic, environmental and toxicological traits: a meta-analysis of 21 years of field data. Sci Rep, 2018. 8(1): p. 3113.
10. Boyle, J.H., H.J. Dalgleish, and J.R. Puzey, Monarch butterfly and milkweed declines substantially predate the use of genetically modified crops. Proceedings of the National Academy of Sciences, 2019: p. 201811437.
11. Wraight, C.L., et al., Absence of toxicity of <em>Bacillus thuringiensis</em> pollen to black swallowtails under field conditions. Proceedings of the National Academy of Sciences, 2000. 97(14): p. 7700–7703.
12. Naranjo, S.E., Impacts of Bt crops on non-target invertebrates and insecticide use patterns. CAB reviews: Perspectives in agriculture, veterinary science, nutrition and natural resources, 2009. v. 4: p. 2009 v.4.
13. Latham, J.R., M. Love, and A. Hilbeck, The distinct properties of natural and GM cry insecticidal proteins. Biotechnol Genet Eng Rev, 2017. 33(1): p. 62–96.
14. Zdziarski, I.M., J.A. Carman, and J.W. Edwards, Histopathological Investigation of the Stomach of Rats Fed a 60% Genetically Modified Corn Diet. Food and Nutrition Sciences, 2018. Vol.09№06: p. 32.
15. Coumoul, X., et al., The GMO90+ project: absence of evidence for biologically meaningful effects of genetically modified maize based-diets on Wistar rats after 6-months feeding comparative trial. Toxicological Sciences, 2018: p. kfy298-kfy298.
16. Zeljenková, D., et al., One-year oral toxicity study on a genetically modified maize MON810 variety in Wistar Han RCC rats (EU 7th Framework Programme project GRACE). Archives of Toxicology, 2016. 90(10): p. 2531–2562.
17. Zeljenková, D., et al., Ninety-day oral toxicity studies on two genetically modified maize MON810 varieties in Wistar Han RCC rats (EU 7th Framework Programme project GRACE). Archives of Toxicology, 2014. 88(12): p. 2289–2314.
18. Young, A.E. and A.L. Van Eenennaam, Prevalence and impacts of genetically engineered feedstuffs on livestock populations1. Journal of Animal Science, 2014. 92(10): p. 4255–4278.
19. Ali, I., S. Zhang, and J.-J. Cui, Bio-safety evaluation of Cry1Ac, Cry2Ab, Cry1Ca, Cry1F and Vip3Aa on Harmonia axyridis larvae. Journal of Applied Entomology, 2017. 141(1–2): p. 53–60.
20. Li, Y., et al., Use of an artificial diet system to study the toxicity of gut-active insecticidal compounds on larvae of the green lacewing Chrysoperla sinica. Biological Control, 2014. 69: p. 45–51.
21. Ali, I., et al., Artificial diet development and its effect on the reproductive performances of Propylea japonica and Harmonia axyridis. Journal of Asia-Pacific Entomology, 2016. 19(2): p. 289–293.
22. King, E.G., J.V. Bell, and D.F. Martin, Control of the bacterium Serratia marcescens in an insect host-parasite rearing program. Journal of Invertebrate Pathology, 1975. 26(1): p. 35–40.
23. Conko, G., et al., A risk-based approach to the regulation of genetically engineered organisms. Nature Biotechnology, 2016. 34: p. 493.