The Sound of Unsound Science: EPA and EU Regulation of Bt Crops and Effects on Non-Target Insects

Greenpeace International, May 2000

Genetically engineered (GE) Bt crops, which include potatoes, cotton and corn, have been a subject of considerable scientific concern about potential ecological risks. After the crops which have been engineered to produce pesticides were approved for the market in the US and Europe several scientific studies have shown adverse effects to non-target species which include e.g. monarch butterfly and lacewing.

Why weren't these effects discovered prior to the approval?

A critical look at the studies submitted by the biotech industry to the US Environmental Protection Agency (EPA) and European authorities shows that the industry studies were mostly inadequate to asses the risks to non-target organisms and the environment. Many of the studies seem to have been designed so that no negative effects could possibly be found, or were terminated before adverse effects would be observable. Even though some of the studies did show adverse effects EPA did not conduct any further studies to determine the effects.

This lack of proper evaluation was reinforced in Europe where the competent authorities in EU Member States and in the European Commission accepted the same studies at face value. Under EU legislation (Directive 90/220/EEC) companies applying for market approval have to submit scientific evidence that the GMOs are safe for the environment. The competent authorities have to examine this information for compliance with the legislation by giving special attention to the environmental risks assessment.

An independent Swiss environmental consultancy EcoStrat on a commission from Greenpeace reviewed 28 separate studies on Bt crops and non-target insects some of these studies were submitted by Novartis for regulatory approval. The report "Review on Non-Target Organisms and Transgenic Bt Plants" by Angelika Hilbeck, Matthias S. Meier and Andrea Raps, also looks at several studies of Bt crops from peer-reviewed publications.

Due to the long persistence of Bt toxin in the field, subtle inter-seasonal and intergenerational effects can be expected, and could result in cumulative effects on biodiversity over a number of years. However the field studies reviewed failed to consider these effects. Of the fourteen field studies twelve looked at just one season, and thus it is not possible to draw conclusive evidence of the effects on non-target species from them. Also, four not only looked at just one field for one season at each location but on very small plots of less than .005 hectares. Further, almost all of the trials identified and listed the species studied at only a rather crude level (order, family) and failed to investigate interaction between natural enemies and prey/host dynamics.

Of all laboratory studies, eight had unreliable conclusions due to highly questionable methods. In five studies adverse effects on non-target insects, lacewings, springtails and ladybirds, were observed.

Of the four feeding studies on non-target insects submitted by Novartis for regulatory approval :

• Only one addressed long-term exposure to Bt. It showed adverse effects on non-target springtails.
• In two it is unclear whether the target species consumed the Bt toxin at all.
• None of the studies appropriately simulated the insect's natural feeding.
• None of the tests considered food-chain interactions or other plant interaction with the insects. feeding on them and the interactions among herbivores and their natural insect enemies.
• All four are modelled on tests used to determine the toxicity of industrial chemicals- none were designed to assess the risks of releasing living organisms into the environment.

Water fleas (Daphnia) are known to be sensitive to harmful metals, and are used to assess the effect of chemicals on water quality. But such a test seems useless for assessing the effects of Bt toxin produced by engineered plants. Pollen and Bt protein dissolves incompletely, if at all, in water. Even if it dissolves, the pollen is enormous compared to the size of food particles that water fleas are able to digest. Therefore, it is most probable that the water flea did not consume any Bt pollen. Nevertheless, the EPA accepted this study.

In this same study which aimed at assessing the effect of Bt on this insect exposed the water fleas for 48 hours. In a chemical test, exposure for a short time would be adequate. But in this case, the 48-hour exposure time renders the test meaningless. Measurement of the effects of Bt toxin are often not reliable until after 48 hours, as even the most susceptible insect larvae don't die immediately after consumption of Bt - exposure effects are generally observed only after a minimum 24 to 48 hours. Moreover, Bt plants produce the toxin continuously, so such a short-term exposure does not simulate field conditions.

Similarly, other tests were done on earthworms (Eisenia foetida) were at best an assessment of the contact toxicity, at worst an exercise in futility. It was not clear that the earthworms in this test were ingesting the Bt material.

A test on honeybee larvae (Apis mellifera L) used pollen that was not pre-digested, as larvae would naturally eat, and exposed the bees to the Bt toxin for just 45 minutes. Such a short-term test of a chemical might be appropriate. But bees near cornfields would be exposed to Bt toxin repeatedly, and throughout their entire life cycles.

Finally, springtails (F. candida) were used in an experimental set-up that is used to assess chronic effects of chemical pesticides. While this test also used questionable feeding material, in this one and only longer-term test, some adverse effects were observed.

EPA fails to react on negative findings

EPA claims that "If adverse effects or potential adverse effects have been observed in the testing, a second or higher tiers of testing has been required to allow EPA to evaluate the risks."

A study Bt-potato fed Colorado potato beetle larvae were provided to ladybird (C. maculata) larvae. Despite the partly inappropriate experimental set-up and the questionable evaluation of the data, this study did reveal some adverse effects of Bt-containing prey on consumption rate and mortality of the ladybird, C. maculata.

In a field study of Bt potatoes, the significant decline of a specialised ground beetle predator (Lebia grandis) was reported because it's primary prey species, the Colorado potato beetle, was eliminated due to the Bt toxin. In the same study, the authors reported a significant increase in the number of leafhopper (Empoasca fabae) adults and nymphs in both the pure Bt-potato fields and the mixed field consisting of both GE and conventional potatoes. Both findings point to important implications of large-scale production of insecticidal GE plants: decline in specialised predators and emergence of secondary pests which can exploit the now abundant food source after the target pest is eliminated.

The study further showed that GE potato growers would still need insecticides, possibly less frequently and targeting another pest, as the remaining predator insects could not be relied on to control remaining secondary pests. Since the target pest is eliminated, it may seem that the decline of a specialised predator is ecologically irrelevant. But the long-term implications of an area-wide decline of such a predator for other cropping systems or natural habitats may still be multifold and should not be dismissed readily.

Despite these adverse effects and potential long-term problems, EPA did not require further, more stringent testing.

According to EPA, Bt crops are defined as "plant pesticides," and thus the focus of regulation is not the living plant, but the pesticidal substance it produces. The assumption underlying EPA's regulation, then, is that an environmental assessment simply needs to look at the expressed Bt toxin as one would look at a Bt spray insecticide. But Bt spray insecticides are different to Bt plants: they contain an inactive cristall-Bt toxin, degrade in as little as a few days, and can be applied in a targeted, rather than broadcast fashion.

Bt crops may produce a pre-activated toxin throughout the life of the plant, at high doses, across an entire field. These differences should trigger requirements for new assessment methods that mimic the new routes of exposure and longer exposure times. EPA says that it regulates "plant pesticides" if there are "new exposures" in the diet or the environment. Yet by relying on these tests, the agency is emphasising false similarities between chemicals and engineered organisms. By ignoring the differences between chemical pesticides and living, plant pesticides, the agency allows industry's faulty scientific approach to go unchallenged.

The EcoStrat report gives strong evidence for regulatory negligence of duty and irregularities by allowing the commercial growing of Bt crops in the US and EU.

Greenpeace exposes fatal flaw in science on biotech crops - press release
Download the full EcoStrat report here (pdf)

top/home