Select Committee on European Communities Second Report - Written Evidence



  1. This paper was prepared to react to a widely published allegation that the Bt gene, which is introduced into crops as a source of resistance to some insect pests, also kills useful insects. The allegations stem from research done by a team of scientists from the Swiss Federal Research Station for Agroecology and Agriculture in Zurich, Switzerland under the leadership of Dr Angelika Hilbeck. The subject species of the work was the lacewing Chrysoperla carnea, a common predator insect. The results of this work were published in the scientific journal Environmental Entomology, volume 27, pages 480-487.

  2. The work of Dr Hilbeck is not new. She started work on lacewings and Bt in 1994, and was assisted by Novartis (then Ciba-Geigy) with material of Bt maize. The objective of the study was to find out what happens to a predator insect (in this case the lacewing larva) when it feeds on an insect that has been eating from plants containing Bt genes. The prey species for the lacewings in the study was the European Corn Borer (ECB), which is the target pest of Bt maize, and which is also a common prey for lacewings in the field.

  3. From 1995 on, Dr. Hilbeck started circulating the results of her first experiments, alleging that she saw more mortality among lacewings feeding on ECB that had been raised on Bt maize than among lacewings feeding on ECB that had been raised on non-Bt maize. Her conclusion was that the Bt in the maize was affecting the lacewings. Both company scientists and academic colleagues immediately challenged the results. It was pointed out that ECB larvae <<raised>> on Bt maize were dying or dead by the time the lacewings ate them. The experiment did not take into account that a predator eating dying prey is going to starve by lack of sufficient food, and/or become poisoned by the internally generated toxins in the decomposing prey. The experiment amounts to feeding a dog with rotting meat until he dies.

  4. Proper control treatments were suggested, such as powdering healthy ECB larvae with Bt just before feeding them to the lacewings, to distinguish between the insect food effect and the Bt effect. Dr. Hilbeck ignored these suggestions. She repeated the above experiment, but included a parallel trial with another prey insect (Spodoptera littoralis, which is less sensitive to Bt). The rationale was that in this second trial, the S. littoralis larvae would not be dying and therefore would constitute healthy prey.

  5. The result was that again more lacewing larvae died when fed on prey that had eaten Bt plants than on the controls. This led to the conclusion that Bt affects a non-target insect (lacewings), and it was published in Environmental Entomology.

6. We would like to deliver three comments on this paper:

    (a)  An assessment of the experimental work done, and the conclusions reached.

    (b)  An assessment of the way in which it was published.

    (c)  An assessment of the wider context into which these results, and other evaluations on the effects of Bt are to be seen.


    —  It is still not understood why the team did not do the obvious control experiment of feeding the lacewing larvae healthy prey powdered with Bt.

    —  In this study, it was assumed that Spodoptera Littoralis was a good control insect, because Dr Hilbeck state that it is insensitive to Bt. It is actually well known that S Littoralis is sensitive to Bt albeit less than ECB. It is quite clear why she still chose this insect as a control, instead of using another leaf eating insect prey of the lacewing.

    —  There have been several other similar studies to investigate possible effects on non-target species. Until now these have come out without showing any effect. Dr Hilbeck mentions some of these studied in passing in the introduction of her paper, but without mentioning which insects were tested. This is significant, because at least two of the studies are about lacewings (Pilcher et al.: Environmental Entomology; volume 26; pages 446-454; 1997) (Sims et al.: Southwest Entomology; volume 20, pages 493-500; 1995), and they reach the opposite conclusion of the Hilbeck study. Dr Hilbeck is aware of this, as she mentions the papers in her references, but in her text, she refers to the work of Pilcher and of Sims without mentioning that his work was on the same lacewing. Therefore, she also does not attempt to discuss why she reaches the opposite conclusion of Pilcher or Sims. In scientific publishing, this is unusual. More unusual still is the fact that the referees of the paper have apparently not picked this up and asked her to address this question. It is unlikely that the referees were unaware of Pilcher's work, since it had been published in the same journal as Hilbeck's work, less than one year earlier.

    —  Dr Hilbeck claims that her work is different from previously published studies in that it tries to study effects of long term feeding. This is correct, but then it obviously becomes crucially important to ensure that the target insects do not die or suffer from the poor health of their prey.

    —  Dr Hilbeck concludes that the lacewings suffer from a Bt related effect. In a way this is correct. If lacewings are fed during their whole development on prey that is dying or diseased, then they suffer as well. In nature, lacewings are non-specialised predators, eating a wide range of insects. They always have healthy prey to eat, and would likely be totally unaffected by the sick ECB larvae. This is actually the conclusion of field observations of Pilcher et al.


  This study has an unusual publication history. Although the experiments were done from 1994 to 1996, it was sent to Environmental Entomology on 28 July 1997, and accepted on 25 November 1997. It was published in April 1998.

  The results of the paper were widely circulated unofficially after the paper was accepted in November 1997. It was cited and commented upon in New Scientist and many other general public media, probably with limited access to the actual paper. From then on, the story of the lacewings started leading its own life, with everyone using it without anyone seeing the actual study. Since the paper itself has only been officially published in April, it is only now that the professional community of entomologists can comment on it.

  In the meantime, the "lacewing problem" has become part of the common wisdom on insect resistant crops, and it is highly unlikely that later criticisms on the quality of the study will reach the same audience as the original study did.


  Bt has been fed to a very wide range of insects, both as a spray and through genetically modified plants. The very fact that the only effect ever noticed is in the above study, which can be heavily criticised for its methodology, is an indication that major effects are highly unlikely. It is impossible to prove this, as it is impossible to prove a negative. What is possible though is to compare Bt plants with other practices to control insect pests, and to do a comparative analysis of damage to non-target insects. This has never been done.

  Companies develop insect resistance through genetic engineering as an alternative to the use of chemical insecticides. In the furore about the lacewing data, it has been largely overlooked that in the standard treatment of ECB today, most or all lacewings in the field are killed by the treatment. The use of Bt genes is a massive improvement over these older technologies in terms of specificity. Until now nothing indicates that lacewings or any other predator insects suffer from the use of Bt maize. But even if there were to be found a small effect, it is already clear that this effect would always be much less than anything we are doing today for control of insect pests.

previous page contents next page

House of Lords home page Parliament home page House of Commons home page search page enquiries

© Parliamentary copyright 1999