D. G Lindsay, CEBAS-CSIC, Murcia
12.3 Assessing the impact of genetically modified crops
12.3.1 Impact on agricultural systems
Genetic modification can have a range of impacts on agricultural systems and therefore will require specific agronomic management. The use of GM varieties would affect the nature of crop volunteers in subsequent crops and require alterations in volunteer management practices. The GM trait may also have an impact if it disperses to other crops and weeds through cross-pollination and seed dispersal. Low-impact genes such as herbicide tolerance, which have little impact on natural environments, become highly significant because of the changes in the herbicide usage required for their management. These herbicides will differ in the effect they have on plant and other species diversity in cropped fields.
Deployment of high-impact genes such as those encoding pest and disease resistance will result in reductions and changes in pesticide usage and thus offer opportunities to enhance diversity in cropped fields, especially if the transgene products are very specific to selected pests. However, it is important that the selection pressures they impose on pests and diseases do not encourage the development of virulent races of pests and pathogens and appropriate management systems are required in order to maintain durable resistance in the GM varieties. Long-term studies on the performance of insect and herbicide- resistant transgenic crops, such as oilseed rape, potato, maize and sugar beet, grown in 12 different habitats and monitored over a period of ten years, showed that no genetically modified plants were more invasive or more persistent than their conventional counterparts.
12.3.2 Impact on uncultivated flora
Genetically modified crops may also have impacts on uncultivated and
‘natural’ environments. These environments may be affected by charac- teristics of crop and wild species induced by novel genetic constructs and their products. Risk assessments must therefore concentrate on whether the genetically modified characteristics of a GM crop and of similarly modified hybridising wild relatives are likely to change the behaviour of the plants or dependent flora and fauna in their environment, to the extent that ecological balances are altered.
12.3.3 Impact on insects and animals
The first successful example of using a foreign plant gene to confer resistance to insects was reported in 1987 (Hilderet al. 1987) and involved transformation of tobacco (Nicotianum tabacum) with the cow pea trypsin inhibitor (CpTi) gene.
Since then there have been many reports of success in insect management using transgenic crop varieties.
The bacterial endotoxins isolated fromBacillus thuringiensis (Bt), comprise one of several groups of proteins which have been shown to have insecticidal properties to a range of economically important insects. Transgenic crop varieties engineered withBtresistance are already in commercial use in the USA and China, while a number of plant proteins, such as inhibitors of proteases, lectins and other digestive enzymes, are being evaluated for their efficacy as insect-resistance mechanisms (Gatehouseet al. 1998).
It is important that genes selected for the control of insect pests have acceptably little effect on non-target insects including predators of the target pest insects, in order to maintain insect diversity in GM crops. Clearly, if there is an effect upon predators that is comparable with current control practices then little benefit will accrue from the deployment of GM crops, a point made strongly by the Royal Society for the Protection of Birds in their submissions on GMOs (1997). Impact assessments are therefore required to examine the effects on non-target organisms in the crop environment.
Studies to evaluate the effect of transgenic plants expressing insect resistance on non-target species have provided, at best, equivocal and often controversial results which have served only to fuel the GM debate rather than provide hard scientific facts on which to base a thorough impact assessment.
Research into the impact of potato plants expressing the snowdrop lectin GNA upon 2-spot ladybirds which feed on the aphid Mysus persicae demonstrated that the ladybirds were affected adversely in terms of fecundity, egg viability and longevity (Birchet al. 1998). However, the authors point out that the effects may either be a direct result of ladybirds preying on aphids which have digested transgenic plant material containing the lectin, or may also be due to poor nutritional quality of the aphids themselves as a food source. Other studies involving the parasitic waspEulophus pennicornisand the tomato moth Lacanobiua oleracea demonstrated that the parasitic wasp was not affected
when it parasitised moth larvae reared on transgenic potato plants expressing the snowdrop lectin GNA (Gatehouseet al. 1997).
More recently Loseyet al. (1999) published a report indicating that pollen from transgenicBt-resistant maize plants had a detrimental effect on the larvae of the non-target Monarch butterfly (Danaus plexippus), which is considered to be a sensitive indicator of environmental disturbance in the USA. Larvae, which normally feed on the leaves of the milkweed (Asclepias curassavica) plant were fed on leaves that had been dusted with unquantified amounts of pollen from the transgenicBtmaize plants. Results indicated that larval survival rate was only 56% compared to 100% survival for larvae fed on leaves dusted with untransformed pollen. Superficially these results indicate an unacceptable environmental impact fromBtmaize. However, closer analyses have revealed a number of serious criticisms of the research, including the use of laboratory studies only, no-choice feeding regimes, lack of stringency, lack of quantification and the use of inappropriate controls (Hodgson 1999). The experiments were not conducted in the field so noin vivodata were available to confirm that (a) milkweeds occur in maize fields, and (b) that Monarch butterflies occur on these milkweeds bearing in mind the insecticide programme received by conventional maize. This once again reiterates the requirement for comprehensive risk assessments based on thorough science.
Schuleret al. (1999) have conducted research concerning the environmental effects of Bt-resistant GM oilseed rape on a non-target insect. The results demonstrated that the behaviour of non-target insects can also play a part in determining howBtplants will affect their populations and should be considered when trying to evaluate the environmental impact of GM crops. Their laboratory-based experiments evaluated the ecological impact of the GM crop on the diamondback moth (Plutella xylostella), a pest that damages the oilseed rape crop, as well as the natural bio-control agent of the diamondback moth, a parasitic wasp (Cotesia plutellae), which kills the moths’ caterpillars by laying its eggs in them. Results demonstrated that parasitoid wasp larvae that were oviposited inBt-susceptible moth larvae not surprisingly died with their hosts. In contrast wasp larvae that had been oviposited inBt-resistant moth larvae feeding on transgenic plants survived and demonstrated no adverse effects of exposure to theBttoxins either as adults or in the development of their own larvae.
The research group then examined the behaviour of the female parasitic wasps in the presence of GM and non-GM leaves. It is known that the female wasps locate the host diamondback moth larvae using herbivore-induced volatiles released from the damaged plants. A wind-tunnel was used to compare the flight response of the wasp towardsBt-susceptible andBt-resistant diamondback larvae which were allowed to feed onBt leaves. The flight and feeding behaviour of each wasp was then measured. In this test, 79% of the parasitoids flew to theBt leaves damaged by resistant moth larvae, with only 21% choosing Bt leaves damaged by susceptible larvae. The apparent lack of effect on the survival or host-seeking ability of the parasitic wasp suggested thatBtplants may have an environmental advantage over broad-spectrum insecticides.
12.3.4 Impact on human health
The potential for transferring genes from one unrelated species to another has caused concern that allergenicity may be introduced into a food source that was previously non-allergenic. An obvious example is that of the recent research programme by Pioneer Seeds where soya bean transformed with a gene from the Brazil nut was found to have allergenic properties. The research programme was halted before any field trials took place but public concern was heightened by reports of this work. All GM foods are now routinely tested for allergenicity using serological tests involving immuno-globulin antigens for specific allergenic proteins.
Transgene instability may be an important issue in the case of transgenic plants engineered to remove the synthesis of harmful toxins. In this situation suppression of gene expression arising from gene flow leading to multiple transgene insertions could prove a serious human or animal health problem if undetected (see Section 12.2.5).
The inappropriate choice of transgenes for achieving a desired trait may also have a serious impact on human health without adequate risk assessment.
Lectins are a group of proteins known to have insecticidal properties that make them attractive candidates for the development of transgenic plants with resistance to the Homoptera insects. They are thought to work by binding carbohydrate side chains present in the gut wall resulting in inhibition of food absorption. As there is the potential of toxicity to humans, it is essential that extensive risk evaluation is required to establish any potential threats of toxicity.
The need for such risk assessment is reflected in work on GM tomatoes (Notebornet al. 1995) and was recently highlighted by reports from the Rowett Institute in Scotland, which indicated adverse immunological and nutritional effects from enhanced lecithin in GM potatoes (Ewen and Pusztai 1999, Fenton et al. 1999). However, aspects of the former of these reports in particular were strongly criticised by a number of reputable scientific bodies as being unsubstantiated and have highlighted the need for agreed methodology in this field of research so that conclusive results can be acquired (Kuiperet al. 1999).
There has also been concern about genetically modified ingredients containing antibiotic resistance genes used to select transformed cells prior to the regeneration of transgenic plants. Use of these genes raises the potential for antibiotic-resistant strains of bacteria to develop via horizontal gene transfer in the gastro-intestinal tract of animals or even humans (Harding and Harris 1997).
This possibility is not thought to be a major hazard since the antibiotic-resistant genes most often used for plant transformation themselves come from bacteria.
They encode resistance against antibiotics rarely used in medicine such as kanamycin, against which a large percentage of gut microflora is already resistant.
However, given the risk, however small, of producing more antibiotic- resistant bacteria, techniques are being developed that will enable selectable markers to be removed from crop plants after the transformation process.
Alternative selectable markers, not based on antibiotic selection, are also being
tested, for example a mannose permease that allows the use of mannose, a sugar not normally available to plant metabolism, as a carbon source during plant regeneration.
Risk assessment methodology will also have to be adjusted for food plants which are modified to improve nutritional and other qualities, a major area for current and future research. Target traits include, for example, improving the nutritional value of proteins, increasing the concentrations of oils low in saturated fats, or fortification with micronutrients or antioxidants. Food plants modified in this way must undergo extensive toxicological and nutritional assessment with a combination ofin vitroandin vivotests, as currently required for all novel foods by the EU, for example. In the case of genetic modification, however, particular attention needs to be given to the detection and characterisation of potential unintended effects of modification. Inferences about such effects can no longer be based solely on chemical analysis of single macronutrients and micronutrients, and known crop-specific antinutrients or toxins. New methods have been developed to screen for potential alterations in the metabolism of the modified organism by such methods as:
• analysis of gene expression (monitored, for example, by microarray technology or mRNA fingerprinting)
• overall protein analysis (proteomics)
• secondary metabolite profiling.
Studies using these designs will need to be designed carefully to take account of the complexity of foods (OECD 1996, Noteborn et al. 2000, Van Hal et al.
2000).