Appendix 6 Genetically modified crops
The shape and structure of agriculture and rural communities has changed significantly over the past century. The term 'Green Revolution' is often used to describe the global transformation of agriculture that has led to significant increases in agricultural production and food security. Drivers of this change include productivity growth through changes to farm practice, global food security and the preservation of biodiversity and the environment, the increasing importance of national and global markets, and the rising influence of consumers on agricultural production. The introduction of genetically modified (GM) crops has been faced with many challenges associated with each of these drivers.
This appendix reviews the extent of GM crop adoption globally, the economic impact of adoption, and issues concerning human and animal health and safety, and the environment. This appendix does not deal with economic or social issues as these have been discussed in the chapters of this report.
A6.1 Extent of the commercialisation of GM crops
Genetically modified crops have been grown commercially in some parts of the world since 1996. The International Service for the Acquisition of Agri-biotech Applications (ISAAA) produces a report annually that characterises the global status of commercialised GM crops. This report has become widely accepted by industry, governments and academics as the authoritative reference for the global deployment of GM crops.
The ISAAA 2006 report states that 10.6 million farmers in 22 countries (11 developing and 11 developed) planted 102 million hectares of GM crops (table A6.1, James 2006). It also states that six countries in the European Union (EU) are now planting commercial GM crops, predominantly insect-tolerant maize. The global area planted to GM crops has increased more than 60-fold over the first 11 years of commercialisation, reputedly making GM crops the fastest adopted crop technology in recent history (James 2006).
The United States of America (US) is the major adopter of GM crops (54.6 million hectares, 53 per cent). In the US, 73 GM events have been approved (deregulated) for commercial cultivation. These events are combinations of 16 different crops and eight different trait or trait combinations. Table A6.2 outlines GM crops commercialised or approved for commercial release globally.
Worldwide, only four crops are commercialised extensively (greater than 50 000 hectares) with GM soybean the principal crop occupying 58.6 million hectares in 2006, 57 per cent of the global GM crop area (James 2006). This was followed by maize (25.2 million hectares, 25 per cent), cotton (13.4 million hectares, 13 per cent) and canola (4.8 million hectares, 5 per cent).
Herbicide tolerance is the predominant GM trait occupying 69.9 million hectares (68 per cent of the global GM crop area), followed by Bt  insect resistance (19 million hectares, 19 per cent) and combined (stacked) traits (herbicide tolerance and insect resistance) occupying 13.1 million hectares (13 per cent of the global GM crop area).
Table A6.1 Countries that have commercialised GM crops
|Argentina||18||Soybean, Maize, Cotton|
|Canada||6.1||Canola, Soybean, Maize|
|South Africa||1.4||Maize, Soybean, Cotton|
|United States||54.6||Maize, Soybean, Cotton, Canola, Squash, Papaya, Alfalfa|
|Total||102 (250 million acres)|
Adapted from James (2006)
There are several important agronomic traits under development in canola and other commodity crops (see reviews by Glover et al. 2005; Holtzapffel et al. 2007). To date, commercialised GM crops have either herbicide, insect and disease resistance traits or combinations of herbicide and insect resistance. GM crops that offer significant promise for sustainable agriculture, and more broadly potential benefits for Australia, include tolerance to environmental stresses (such as drought, frost, salinity and acid soil tolerance), pests and disease for example virus resistance in white clover; cane grubs in sugarcane) and human and animal health benefits (such as speciality oils with improved quality and quantity and composition). Some of the current and future GM based products in canola are described in box A6.1.
Table A6.2 GM crops commercialised or approved (deregulated) for commercial release globally
Control / HT
This table provides details of GM crops that have been commercialised globally and GM crops that have been approved for commercial cultivation and sale (deregulated) by the US Department of Agriculture (September, 2007). GM crops are grouped by trait or trait combination (see Legend below). Boxes with an 'A' represent GM crops that are commercially grown around the world. Boxes with a 'B' represent GM crops that have been deregulated in the US but are not yet commercially grown. An empty box signifies that there are no deregulated trait/crop combinations in the US. The Box with a 'C' (alfalfa) was initially deregulated and subsequently returned to regulated status in March 2007.
Legend: HT = herbicide-tolerant; IR = insect-resistant; VR = virus-resistant; HT/IR, IR/VR and pollination control/HR= 'stacked' crops with both of the indicated traits.
Pollination control is used for breeding purposes and hybrid vigour. Quality traits include altered oil composition (soybeans and canola), altered protein composition (corn), reduced nicotine levels (tobacco), flower colour (carnations) and altered fruit ripening (tomato). Note that in some cases a letter in a box represents several GM crop 'events' - or different types of the same basic crop-trait or combination. Data based on James (2006) and USDA data from www.aphis.usda.gov/brs/not_reg.html (accessed September, 2007)
Part of the GM debate centres on the extent of the commercialisation and why it is currently limited to only four major crops (maize, soybean, cotton andcanola) and a small number of traits. This is in contrast to the large numberof benefits that potential GM technologies could have, including human
health and climate change. Possible reasons for the delay or withdrawal from the market of several deregulated GM events include commercial decisions, marginal consumer acceptance (for example, Flavor Savr™ tomato, Kramer and Redenbaugh 1994) and the financial cost of a pathway to market (Kalaitzandonakes
et al. 2007). Of these the cost of regulation to meet legislative requirements for the approval of GM crops is often prohibitive. The
cost of deregulation for a single GM event (that is, satisfying all regulatory requirements to gain approval for commercial release) is estimated at $US6–20 million and depends on the trait, crop and intended market(s) (Kalaitzandonakes et al. 2007).
Box A6.1 Current and future GM canola traits
Australia - commercial production
- Roundup Ready® GM canola. Tolerant to the herbicide glyphosate. Developed by Monsanto and approved for commercial release in Australia in 2003.
- InVigor® GM canola. Tolerant to the herbicide glufosinate ammonium and also two genes that control sterility and fertility in pollen to enable hybrid breeding. Developed by Bayer CropScience and approved for commercial release in Australia in 2003.
Australia - research and development:
- Bayer CropScience has received approval to trial GM Canola crops in Victoria and South Australia, which are tolerant to a novel herbicide (DIR 032/2002).
- The Commonwealth Scientific and Industrial Research Organisation (CSIRO) have developed GM Brassica napus and B. juncea with higher oleic acid levels (up to 89 and 73 per cent respectively). This is yet to be field trialled.
- CSIRO have examined the synthesis of long chain polyunsaturated fatty acids (LC PUFAs), mainly Omega-3 and Omega-6 in plants.
Overseas - commercial production
- Roundup Ready® GM canola is currently grown commercially in Canada. In 2005, this represented about 48 per cent of the canola grown in Canada.
- LibertyLink® GM canola (tolerant to the herbicide glufosinate ammonium) is currently being commercially grown in Canada. In 2005, this represented about 34 per cent of the canola grown in Canada.
- GM bromoxynil herbicide-tolerant canola was commercialised in Canada during 1998, however its production has now been discontinued.
Overseas - research and development
- In the US, a field trial of GM Lepidopteran (caterpillar) resistant canola is currently underway. Other GM canola traits at different stages of research and development in the US and European Union include: Increased nitrogen use efficiency, increased oil content, increased tolerance to environmental stress (e.g. drought tolerance), increased seed size, and reduced pod shatter.
- In the European Union, a major study was conducted from 1999–2005 to improve the nutritional value of B. napus. The study investigated oilseed compounds such as tocopherols and lecithin, and the benefits of introducing compounds such as resveratrol (an antioxidant) and LC PUFAs.
- Monsanto have developed a GM canola crop (Laurical®), which produces high levels of lauric acid and can be used as a substitute for coconut and palm kernel oil.
- Research has been undertaken to produce GM canola varieties with increased lysine levels.
- The use of GM canola meals with altered phosphorous, phytic acid and sinapine levels could improve the nutritional value of canola meal as stockfeed for animals. Research has already been undertaken on GM soybean and the technology could be transferred to canola.
- Canola variety with increased lauric acid and myristic acid content, which can be used in detergents - this is now at a commercial stage of development.
- Canola variety with high stearic acid content, which can be used in grease - this is yet to be commercially produced although development is complete.
- A canola variety with petroselenic acid, which can be used in food - this is still under development.
- A rapeseed variety in Canada with increased erucic acid, which can be used in plastics and lubricants is under development.
- Research is being undertaken by Metabolix to produce a plastic called Polyhydroxybutyrate in GM canola crops.
Source: data from Glover et al. 2005 and Holtzapffel et al. 2007
It is worth noting that a lot of gene technology research is not intended to lead to direct commercial outputs. Further, whilst many public sector research and development programs are clearly focused on commercialisation, a number of obstacles often prevent a path to market such as technical barriers, freedom to operate, regulation, value capture, marketing uncertainty and the costs of deregulation (see Glover et al. 2005).
In Australia, GM cotton tolerant to herbicides and/or resistant to insects is the only broadacre GM crop grown by farmers. In 2005-06, more than 840 000 tonnes of Australian cotton were produced over 335 000 hectares in New South Wales and Queensland (Australian Bureau of Statistics 2006).The Australian cotton industry is valued at approximately $1.5 billion, with GM varieties making up 90 per cent of all cotton grown. This is the maximum proportion of GM cotton that can be grown, given the isolation and trap crops required for GM cotton production as part of sustainable pest management. Apart from cotton lint, the uses of GM cotton seed almost mirror those of non- GM canola: GM cotton seed protein meal is used in the animal feedlot industry and the dairy industry, and as drought feed for cattle, while cotton seed oil is used extensively in commercial cooking in Australia.
A6.2 Economic impact of the commercialisation of GM crops
Globally, total agricultural production has significantly increased over the past 40 years, with the greatest increases (in percentage terms) in developing countries (Mattson and Koo 2006). Investment in research and development in agriculture increases productivity to not only contribute to food security but also to help countries to maintain and improve international competitiveness in global markets. Investment in biotechnology is providing significant impacts on productivity and competitiveness as well as opening new non-food use markets for agricultural products such as biofuels and bio-based products (Mewett et al. 2007).
Grower adoption of new technologies and practices, including GM, is primarily driven by economic considerations but increasingly include an expectation of both tangible and intangible benefits. In general, the impact on and off-farm in GM adopting countries is reportedly positive and derived from enhanced productivity, quality and improvements to efficiency (Gómez-Barbero and Rodríguez-Cerezo 2006).
Since the commercialisation of GM crops, reports indicate that global farm incomes have benefited by $US24.2 billion with 47 per cent of this benefit accruing to farmers in developing countries (Brookes and Barfoot 2006a). In Europe, farmers who have adopted GM maize (Bt insect and/or herbicidetolerant) have, on average, earned additional annual income levels of between €65 and €141 per hectare, equal to increased profitability of 12–21 per cent (Brookes 2007).
In the three years to 2005-06, more than 60 per cent of the world trade in soybeans, maize, canola and cotton came from countries that grew GM varieties (US Department of Agriculture 2007). In the case of soybeans, world trade is dominated by GM varieties, with only 8 per cent of trade in certified non-GM soybeans (Foster and French 2007).
A number of studies have examined the economic benefits of GM crops in the US (Falck-Zepeda et al. 2000 a and b; Fernandez-Cornejo and Caswell 2006; Moschini et al. 2000; Price et al. 2003; Sujatha 2006). A recent comprehensive study by Sujatha (2006) indicates that in 2005, the use of biotechnology contributed to increased US agricultural production by 8.34 billion pounds (3.7 million tonnes), lowered crop production costs by $US1.4 billion dollars, and reduced pesticide use by 69.7 million pounds (31 615 tonnes) on 123 million acres.
The economic benefits of insect resistant cotton in India have been disputed. The first economic study estimated potential yield benefits of up to 80 per cent (Qaim and Zilberman 2003) but focused on farm-level field trial data and did not reflect actual farm experience with commercial cultivation (Raney 2006). A subsequent study considered four different states within India and found significant positive impacts on farmer returns (Qaim et al. 2006). However, there was considerable variation between states, which the authors attribute to the use of varieties not well suited to the growing conditions in those states.
There have been reports of significant direct and indirect social and economic benefits since the introduction of GM crops into Argentina in 1996. A study by Trigo and Cap (2006) assessed the economic impact of the adoption of GM crops. For GM soybeans the total accumulated benefits for 1996–2005, allowing for the replacement of sunflower, cotton and pasture agricultural production, were estimated at $US19.7 billion. They found that the benefit was distributed 77.5 per cent to farmers, 4 per cent to seed suppliers, 5 per cent to herbicide suppliers and 13.5 per cent to the Argentine Government as revenues collected through an export tax that was imposed in 2002. Trigo and Cap (2006) concluded that the release of herbicide-tolerant soybeans may have contributed to the creation of almost one million jobs (whole economy-wide), representing 36 per cent of the total increase in employment over the period 1996–2005.
Intangible economic benefits have also contributed to the rapid adoption of GM crops. These include but are not limited to: increased flexibility in weed and pest management practices, synergistic benefits with zero or reduced tillage, increased quality through reduced mycotoxins and cleaner crops, increased health and safety of farm workers through reduced handling of pesticides and reduced effects of soil-incorporated herbicide residues on subsequent crops (Beckie 2006; Brookes and Barfoot 2006 a and b; Wu 2006).
A6.3 The effects on human and animal health of GM crops
The debate over GM crops, and in particular GM food, has highlighted the potential positive and negative impacts of agriculture on human and animal health. Nutritional and safety assessments of GM foods have featured prominently with extensive study of GM crop nutrition and the fate of novel DNA and proteins in livestock products such as meat, milk and eggs (reviewed by Aumaitre et al. 2002; Flachowsky et al. 2005 a and b; Flachowsky et al. 2007).
The majority of studies have focused on assessment of herbicide-tolerant and/or insect pest resistant GM cotton, canola, wheat, rice, maize, potato and sugar beets or byproducts from these GM crops. Animal feeding studies with GM crops have been conducted with rodents, poultry (laying and broiler), dairy and beef cattle, sheep, pigs, rabbits, fish, and quails (Flachowsky et al. 2005a; Taylor et al. 2007 a and b).
To date, no significant differences in the safety or the nutritional value of GM foods have been observed compared with non-GM varieties. Further, no residues of GM DNA or proteins have been found in any organ or tissue samples obtained from animals fed GM crops (Flachowsky et al. 2005a). Despite these studies showing no differences in comparison to conventional crops, a few studies have suggested potential impacts of GM crops on health, specifically: the use of a Brazil nut gene in soybeans (Nordlee et al. 1996) and potato (Tu et al. 1998); the use of lectin genes in potatoes (Ewen and Pusztai 1999) and field peas (Morton et al. 2000); Bt genes in maize (Hammond et al. 2006; Séralini et al. 2007); and herbicide tolerance genes in soybean (Sakamoto et al. 2007). Each of these studies will be examined individually below.
A6.3.1 Improved nutrition with a Brazil nut gene
Soybean is a major source of protein meal for both humans and animals.
However, nutritional quality is compromised by a relative deficiency in the amino acid methionine. In an attempt to improve nutritional quality, Pioneer Hi-Bred introduced a methionine-rich gene from the Brazil nut into soybeans. Variations of this gene have also been transferred into potatoes (Tu et al. 1998). Both studies demonstrated an improvement in amino acid levels. However, since the Brazil nut is a known allergenic food, the allergenicity of the Brazil nut gene in soybean was also assessed (Nordlee et al. 1996). The study demonstrated that an allergen from a food known to be allergenic can be transferred into another food by genetic modification. Pioneer has since terminated the work and no commercial GM crops have been developed with the Brazil nut genes due to possible allergenicity effects.
A6.3.2 Use of lectin genes for insect pest resistance
One of the most controversial studies examined the effects of GM potatoes on the digestive tract of rats (Ewen and Pusztai 1999). Laboratory rats were fed three different potato diets, plain potato, potato supplemented with snowdrop flower lectin, and GM potato making its own snowdrop lectin. The study claimed appreciable differences between the small intestine and large bowel of rats fed GM potatoes and potatoes supplemented with snowdrop lectin compared with rats fed non-GM potatoes. Interpretation of data in this gene) derived from kidney bean to provide protection against the pea weevil (Morton et al. 2000). Preliminary rat feeding studies with the GM pea established minimal detrimental effects on the nutritional quality of the GM peas compared with non-GM peas (Pusztai et al. 1999). Although, the GM peas were resistant to the pea weevil, the researchers discovered that the addition of GM peas to pig and poultry diets reduced productivity by up to 10 per cent and in a subsequent study when fed in small quantities to mice over a few weeks, it caused mild allergenic inflammation in the lungs (Prescott et al. 2005). The detrimental impacts on pig and poultry productivity were considered too high and the project was discontinued in October 2003.This study demonstrates that expression of non-native proteins in plants may lead to the synthesis of structural variants with altered immunogenicity. Safety assessments conducted by food safety regulators (for example, Food Standards Australia New Zealand) carefully examine any potential toxicity or alllergenicity effects of GM foods before approval for sale.
A6.3.3 Safety of Bt maize MON863
European Union (EU) safety assessments for the GM maize line MON863 have received significant attention. GM line MON863, also commercialised as YieldGard® Rootworm, is genetically modified to express a Bt gene that provides resistance to attack by a larval form of a beetle (cucumber or asparagus beetle). This Bt gene is different from the Bt gene expressed in other currently commercialised GM maize (e.g. MON809, MON810 and Bt11), which are resistant to the European corn borer and Mediterranean stem borer.
In 2002, Monsanto applied to the EU to have MON863 maize and the maize hybrid MON863 x MON810 approved for commercialisation. Following a request from the French competent authority, Monsanto provided data from a 13-week rat feeding study. The GMO Panel of the European Food Safety Authority (EFSA) examined the results of the feeding study and concluded that the results did not indicate adverse human or animal health effects from consumption of MON863. The MON863 and hybrid MON863 x MON810 lines were subsequently approved in 2004 for commercialisation in the European Union. MON863 has also been approved for import and food use in Australia, Japan, Korea, Taiwan, the Philippines, Russia and Mexico.
Following EU approval, Monsanto published the data on the 13-week rat feeding study (Hammond et al. 2006) but were reluctant to make publicly available the raw data of the study. An Appeal Court action brought by Greenpeace forced Monsanto to publicly release the data in 2005.
In March 2007, a report of a new statistical analysis of the Monsanto study was published that raised toxicity concerns in the rats fed MON863 (Séralini et al. 2007). The re-analysis identified statistically significant differences in serum protein values in rats fed an 11 per cent or 33 per cent diet of MON863 and results of this study did not indicate adverse human or animal health effects from consumption of MON863). Similarly, Food Standards Australia New Zealand (FSANZ) conducted an independent assessment of the results of the Monsanto study and came to the conclusion that the data do not indicate any adverse effects, which could be attributed to MON863 corn or its constituents. FSANZ (2004) further stated that observed differences were consistent with normal physiologic variation and do not indicate adverse effects from the consumption of MON863 maize.
Following this study, EFSA re-examined their assessment of MON863, but could not find any reason to change their original conclusions (that is that the results of this study did not indicate adverse human or animal health effects from consumption of MON863). Similarly, Food Standards Australia New Zealand (FSANZ) conducted an independent assessment of the results of the Monsanto study and came to the conclusion that the data do not indicate any adverse effects, which could be attributed to MON863 corn or its constituents. FSANZ (2004) further stated that observed differences were consistent with normal physiologic variation and do not indicate adverse effects from the consumption of MON863 maize.
Speculation remains as to the interpretation of the data provided by Monsanto. However, key regulators such as EFSA and FSANZ are in agreement that variations between animals fed GM maize compared to non-GM maize appear to have occurred randomly, were generally of a small magnitude, and within the normal range for laboratory rats (EFSA 2007; FSANZ 2004). As such, both regulators have approved MON863 for import and food use.
A6.3.4 Reduced mycotoxins in Bt maize
Like most other grains, maize kernels can be infected with fungi before and after harvest. Mycotoxins are toxic compounds produced by several plant fungal pathogens such as Fusarium ear rot and this can affect the nutritional value and safety of maize as a human food or animal feed.
Two of the most agriculturally important mycotoxins are fumonisins and aflatoxins. Fumonisins are found almost exclusively in maize, while aflatoxins are found in a variety of crops including maize, cotton, peanuts, pistachios, almonds, and walnuts (Robens and Cardwell 2003).
Insect damage to maize kernels can often lead to greater levels of Fusarium contamination and hence greater levels of mycotoxin. Field studies of Bt maize have consistently demonstrated that hybrids containing Bt insect resistance have a significantly lower incidence and severity of Fusarium ear rot and subsequently produce grain with lower mycotoxin concentrations than their non-Bt counterparts (Munkvold et al. 1999; Piva et al. 2001). As a consequence of this, Piva et al. (2001) observed significant weight gain in piglets fed Bt maize compared to those fed the non-GM line.
It has been reported that the total benefit of Bt maize's reduction of fumonisin and aflatoxin in the United States was estimated at $US23 million annually (Wu 2006). The report also predicts that the strict marketing standards of these two mycotoxins could result in global trade losses in the hundreds of millions of dollars annually, with the United States, China and Argentina suffering the greatest losses. Therefore, any reduction in the prevalence of these mycotoxins would have a considerable positive impact on global trade.
A6.3.5 Herbicide-tolerant soybeans
Herbicide-tolerant soybean was the largest GM crop grown in 2006 (James 2006). The health and safety of herbicide-tolerant soybeans have been examined using animal feeding studies. These have been conducted over various lengths of time and results have been 'inconsistent', possibly due to differences in experimental conditions.
Research published by an Italian group indicated differences in cellular structure and function, but not gross morphology, in the liver, pancreas and testis of mice fed diets supplemented with GM and non-GM soybean (Malatesta et al. 2002; Malatesta et al. 2003; Malatesta et al. 2005; Vecchio et al. 2003).
In contrast, Brake and Evenson (2004) studied mouse testis as a sensitive model to examine potential toxic effects of GM soybean. Pregnant mice were fed diets containing GM or non-GM soybean through gestation and lactation. After weaning the young male mice were maintained on the respective GM diets. Multi-generation studies were conducted with results indicating that there were no differences between mice fed a GM or non-GM soybean diet and that GM soybean had no negative effects on testicular development in mice.
Sakamoto et al. (2007) also conducted a long term feeding study on rats, examining the influence of GM soybeans compared with that of the non-GM soybeans, and a commercial diet. At 26 and 52 weeks, animals were subjected to extensive biochemical and pathological examination. There were several differences in animal growth, food intake, biochemical parameters and histological findings between the rats fed the GM and/or non-GM soybeans and the rats fed a commercial diet. However, body weight and food intake were similar for the rats fed the GM and non-GM soybeans. Autopsy findings, biochemical parameters, organ weights, and pathological findings showed no differences between rats fed the GM and non-GM soybeans, indicating that the long-term feeding of a GM soybean diet has no apparent adverse effect in rats (Sakamoto et al. 2007).
One of the most controversial GM feeding studies has been the unpublished data provided by Dr Irina Ermakova of the Institute of Higher Nervous Activity and Neurophysiology of the Russian Academy of Sciences. The limited information presented by Ermakova has been widely criticised by a number of authorities by having insufficient detail and not being subjected to peer review. Recently, Marshall (2007) provided an account in Nature Biotechnology of the work appended with comments from researchers in the field, who noted that due to the lack of information concerning the composition and nutritional adequacy of the test diets and the abnormally high incidence of mortality in the control animals, a number of other explanations for the results presented remain open, apart from the effects of the GM and non-GM soybean diet.
A6.4 Environmental impact of GM crops
Increases in agricultural productivity have been directly related to advances in technology such as farm mechanisation, advances in plant and animal breeding, increased yields and quality enhanced by the rapid development of inexpensive chemical fertilisers and pesticides, development of crop rotation and reduced tillage, the propensity to grow single crops over a large area and and gene flow into non-GM crops and related species. Indirect effects might include impacts on pest and weed management (including chemical use), and effects on natural habitats and ecosystems. Through an understanding of these potential impacts, breeding programs and crop management practices may be tailored to ensure environmental safety and future sustainable crop production (for example see review by Beckie 2006).
A6.4.1 Potential effects on non-target organisms
There have been concerns that insect resistant GM crops could harm organisms other than the pests targeted by the insect resistance genes (non-target organisms). Potential effects could be directly associated with the mode of action of the introduced gene(s), or indirectly associated with changes in the availability of and or quality of non-target organisms (reviewed by Glare and O'Callaghan 2000; O'Callaghan et al. 2005; Romeis et al. 2006).
In order to be directly affected, non-target organisms would need to ingest the insecticidal protein(s) from the GM crop. This can occur via ingestion of GM plant tissue, feeding on insects that have consumed the GM crop, or from the environment from the persistence of GM residues in the soil.
Impact on lacewings
Several studies have suggested that Bt maize may have negative impacts on the beneficial insect, the green lacewing (Hilbeck et al. 1998 and 1999). This insect occurs naturally but is also used widely for the biocontrol of many different insect pests including those of non-GM maize. The adult lacewings do not kill insect pests, but their larval offspring are predacious on the eggs and the immature stages of most soft-bodied pests such as aphids, thrips, spider mites, whitefly, mealybugs, leafhoppers, and the eggs and caterpillars of most pest moths. Under laboratory conditions, elevated mortality was observed in lacewing larvae reared on the larvae of Egyptian cotton leaf worm or European corn borer that were fed Bt maize (Hilbeck et al. 1998 and 1999). In another study, effects on lacewings were only observed when fed with Egyptian cotton leaf worm but not when fed with spider mites or aphids reared on Bt maize (Dutton et al. 2002). The authors suggest that the effects observed were related to the quality of the prey rather than as a direct effect of the Bt protein. This conclusion is supported by bioassay studies examining the direct effect of the Bt protein on lacewing larvae (Rodrigo-Simon et al. 2006; Romeis et al. 2004).
Several independent field studies (e.g. Candolfi et al. 2004; de la Poza et al. 2005; Pilcher et al. 2005; Whitehouse et al. 2005) have revealed only minor, transient or inconsistent effects of Bt crops when compared with non-Bt controls (reviewed by Romeis et al. 2006). The main effect appears to be a reduction in specialist predator populations within crops as commonly seen with the use of other pest control methods such as chemical pesticides, biological control agents and the use of crop rotation to reduce the availability of host plants.
A recent comprehensive analysis of 42 field experiments to examine the effect of Bt cotton and corn on non-target invertebrates found that non-target invertebrates are generally more abundant in Bt cotton and Bt corn fields than in non-GM fields managed with pesticides (Marvier et al. 2007). The abundance of non-target insects was less in GM fields compared with insecticide-free control fields.
Impact on honey bees
Many plants rely on insects, such as honey bees, to facilitate cross pollination. The impact of GM crops on honey bees is of both ecological and economic importance. As such, many regulatory risk assessments include studies that directly evaluate the potential toxicity of a GM crop on bees (for example, EPA 2001).
Studies of honey bees fed with Bt proteins and with pollen from Bt crops(cotton or maize) have identified no negative effects on the health or flight activity of honey bees (Babendreier et al. 2004, 2005 and 2007; Hanley et al. 2003; Liu et al. 2005; Malone et al. 2001; O'Callaghan et al. 2005; Sanvido 2006). Similarly, pollen from GM canola resistant to glyphosate had no effects on honey bee health (Huang, Hanley et al. 2004; Mohr and Tebbe 2007) and GM protein has not been detected in the nectar and pollen of different GM canola varieties (Pham-Delègue et al. 2002).
Both of these studies were criticised because, whilst the Bt maize pollen might be toxic to larvae, ecological data such as the spatial and temporal patterns of the butterfly and Bt maize might preclude impacts on butterfly populations. In other words this result may not be ecologically relevant. Additionally, the Bt maize event used for that research (Event 176), expressed significantly higher Bt protein levels in pollen and anthers compared with the most widely planted Bt maize events (MON810 and Bt11) and is no longer commercially available (Stanley-Horn et al. 2001).
Recent studies also suggested that monarch butterfly larvae exposed to the anthers of Bt maize behave differently than if exposed to non-GM anthers and hence may indirectly affect populations (Anderson et al. 2004 and 2005; Prasifka et al. 2007) but the ecological significance or relevance of this finding remains uncertain.
A two-year study in Canada suggests that the impact of pollen from commercially grown Bt maize on monarch butterfly populations is negligible (Sears et al. 2001). Evidence included laboratory and field studies showing no acute toxic effects at any pollen density, limited overlap between pollen and anther distribution and monarch butterfly larval activity, and only a portion of monarch butterfly populations use milkweeds in and around Bt maize fields (Anderson et al. 2004; Sears et al. 2001).
Soil populations (for example, micro-organisms, earthworms, mites and nematodes) are affected by a number of factors and agricultural practices including soil type, level of cultivation, history of the location, level and frequencies of pesticide application and type of crop and variety grown. It has been demonstrated, for example, that soil microbial populations can vary between varieties of the same crop (reviewed by Garbeva et al. 2004). Similarly, numerous studies have investigated aspects of the biology and ecology of soil micro-organisms (including plant pathogens) with respect to herbicide application (reviewed by Altman and Campbell 1977; Altman and Rovira 1989).
The introduction of herbicide resistant GM crops has raised questions of the effects of herbicides on soil micro-organisms and in particular plantpathogens. In a recent field study at six sites in Canada, Lupwayi et al. (2007)evaluated the effect of glyphosate resistant wheat and canola crops on soilmicro-organisms under reduced tillage or conventional tillage systems. Results indicated that the effects on soil microbial biomass, bacterial functional diversity and community structure, and dehydrogenase enzyme activity were minor and inconsistent over the wide range of growing conditions and crop management (Lupwayi et al. 2007).
Several reports have described synergistic effects between the herbicides used on GM crops (for example, glyphosate and glufosinate) and plant pathogens (Ahmad et al. 1995; Desclazo et al. 1998; Fernandez et al. 2005; Kremer et al. 2005). For example, sudden death syndrome (SDS) is a fungal disease of soybeans caused by Fusarium solani f.sp. glycines, and has caused substantial soybean yield reductions in the United States. An increase in the incidence of this plant pathogen was linked to the use of glyphosate in GM soybean and caused widespread concern amongst US farmers (Sanogo et al. 2000). However, a three-year field study evaluating both susceptible and resistant soybean cultivar response to four herbicides found that the response of GM soybeans to SDS was not different from responses by conventional soybeans to herbicide application (Sanogo et al. 2001). Further, the application of glyphosate to SDS resistant soybean cultivars did not alter their resistance and the authors proposed that poor cultivar selection was the most likely cause of SDS in soybeans and that selection of resistant cultivars was the most effective solution to the problem (Sanogo et al. 2001). Similar observations and conclusions have been reported for Sclerotinia stem rot and a damping-off pathogen of soybean (Harikrishnan and Yang 2002; Nelson et al. 2002). This research highlights the importance of complementary roles of gene technology and plant breeding.
Although Bt has been shown to bind to soil particles (for example, see Stotzky 2004), a number of studies have found no evidence of accumulation in the soil after several years of cultivation (for example, Baumgarte and Tebbe 2005; Dubelman et al. 2005). However, CSIRO research examining the ecological impacts of GM cotton on soil biodiversity observed the presence of Bt in the roots of Bt cotton varieties in different soils and that Bt protein was released by cotton roots into solution culture and in soil (Vadakattu and Watson 2004). Thesignificance of bioaccumulation of Bt from GM cotton remains unknown. In field studies with Bt maize compared with non-GM maize, soil micro-organismdiversity appeared to be more impacted, primarily by variation in the soil and the age of the plants grown (Baumgarte and Tebbe 2005) and cultivar variation (Griffiths et al. 2005).
Overall, studies examining the effects of GM crops on soil micro-organisms have been inconsistent (reviewed by Bruinsma et al. 2003; Dunfield and Germida 2004; Motavali et al. 2004). To date, laboratory and field studies have not been able to demonstrate toxicity of GM products to non-target soilorganisms (Sanvido et al. 2006). The major impediments to understandingany potential effects of GM crops on soil communities include the lack ofknowledge of the background variation within agricultural systems and the highly diverse soil environments that occur (Griffiths et al. 2005; Sanvido et al. 2006). Without an understanding of this information, it remains difficult to establish the ecological relevance of any differences that might be observed in soil ecosystems, regardless of whether the crops are GM or non-GM.
Farm-Scale Evaluation experiments in the United Kingdom
In 1999, the UK government asked an independent consortium of researchers to investigate how growing herbicide-tolerant GM crops might affect farmland wildlife compared with growing non-GM varieties of the same crops. The Farm Scale Evaluation (FSE) study was designed to test the hypothesis that there is no difference in biodiversity between GM crops and conventional crops. The study was carried out over three years in 60 fields across England and Scotland. The crops grown were sugar beet (including fodder beet), maize, and winter and spring canola, and the biodiversity recorded included the abundance of weeds and invertebrates. The data were rigorously analysed, peer-reviewed, and published in a series of eight papers in the Philosophical Transactions of the Royal Society of London (reviewed by Freckleton et al. 2003).
Results of these studies indicated that there were reductions of between 60–80 per cent in weed biomass at the end of the growing season in GM sugar beet and spring GM canola, reflecting increased weed control in these crops. In contrast, there was an increase of 82 per cent in weed biomass in GM maize compared with the conventional maize crop. The authors suggest that a reason for this is that pre-emergence weed control using the herbicide atrazine in conventional maize is extremely efficient, and that the herbicide-resistant GM maize system is unable to improve on this.
Additional findings from the study included: difference in biodiversity were more related to the type of crops rather than GM or not; lower herbicide inputs to GM crops and reduced numbers of bees, butterflies and Heteroptera (certain bugs) in GM sugar beet and GM canola due to reduced weed populations (Freckleton et al. 2003).
Following the release of data from this study, the CSIRO conducted an independent appraisal of the results as they might apply to Australia (Lonsdale et al. 2003). Conclusions from their assessment identified that although the FSE study was important and well designed, it has only limited relevance to Australia. For example, the study confirmed that the impacts that GM crops might have on biodiversity, relative to conventional crops, would depend on the crop, the type of genetic modification and the crop management strategies employed (Lonsdale et al. 2003).
A6.4.2 Gene flow issues associated with GM crops
A major concern amongst opponents of GM crops are claims that a GM trait(s) will move into other organisms, non-GM crops and wild relatives, especially those with a propensity to become weeds (Snow 2002).
Gene flow from plants to other organisms
The risk of GM traits transferring to other organisms (horizontal gene transfer) such as humans, animals, fungi, bacteria, other micro-organisms and viruses is reportedly negligible (Flachowsky et al. 2007; Mohr et al. 2007; Nielsen 1998). Since the introduction of GM crops, no genetically modified DNA has been integrated into the cells of animals fed GM crops (Flachowsky et al. 2005a) or insects (Mohr et al. 2007). In general, cross species gene transfers could only be detected over evolutionary time scales (Ochman et al. 2000) and comparison of the sequences of plant and bacterial genes suggests that horizontal gene transfer from plants to bacteria during evolutionary history has been extremely rare, if it has occurred at all (Nielsen 1998). For gene transfer to happen, a complex series of events would be required and studies indicate that natural gene transfer between plants and other organisms is also exceedingly rare (Aoki and Syono 1999; Mayo and Jolly 1991). Gene transfer from plants to bacteria has not been demonstrated under natural conditions (Nielsen et al. 1997; Nielsen 1998; Syvanen 1999) or through deliberate attempts to induce natural transfers (Coghlan 2000; Schlüter et al. 1995). However, the transfer of plant genes to bacteria and viruses has been demonstrated in laboratory and glasshouse experiments between both the plants and bacteria following heavy unnatural selection methods (Greene and Allison 1994; Nielsen et al. 2000; Schoelz and Wintermantel 1993).
Gene flow between plants
Gene transfer between plants through cross pollination or out-crossing can only occur between related and sexually compatible plants that are in very close proximity and that flower synchronously (Glover 2002). Gene flow occurs naturally via pollen from species that are partial or obligate out-crossers (for example, canola and maize) but is less likely from plants that predominantly self pollinate (for example, soybean, wheat and rice). The distance that viable pollen can travel is influenced by the dispersal method (for example, insect or wind dispersal), environmental conditions, the weight and the longevity of the pollen, many of which are species dependent (Rieger et al. 2002).
Pollen movement per se is not the primary issue associated with gene flow. Any potential problems associated with gene flow from GM crops should consider the likelihood of introgression of the GM traits into populations, not just the movement of pollen (reviews by Conner et al. 2003; Hails and Morley 2005; Nap et al. 2003; Stewart et al. 2003). Introgression is dependent on the nature of the gene and the biology and ecology of the recipient plant (Hails 2002). It is also dependent on the gene having a selective advantage, such as herbicide resistance.
Plant seeds provide another important mechanism of gene flow. Seeds from one cultivar can be inadvertently mixed with the seeds from another in agricultural equipment and through bulk handling. Seeds can also remain viable in the soil following harvest for many years, germinate and grow in subsequent crops as a 'volunteer'. Any seed formed from volunteers could then be combined with the seed of a subsequent crop.
It has been proposed that the management of herbicide resistant canola volunteers is best controlled in-crop (Harker et al. 2006). Further, it is reported by Beckie et al. (2004) that with appropriate weed management strategies, allsingle or multiple herbicide resistant volunteers that might form can effectively be managed on-farm using herbicides with alternative modes of action. For example, glyphosate resistant volunteers can be efficiently controlled with the herbicides metribuzin or 2,4-D.
Maize is highly out-crossing but is not known to form feral populations, does not have any weedy relatives and only rarely forms volunteers (Warwick and Stewart 2005). Gene flow from the pollen of Bt maize into non-GM maize has been investigated with the authors concluding that current isolation distances used for GM maize are sufficient to manage potential outcrossing (Ma et al. 2004).
Soybeans are generally not considered to be a serious volunteer weed problem (Owen 2005). Although some closely related soybean species exist, introgression of GM traits has not been reported either due to the lack of closely related species growing in soybean production areas or they do not have a competitive advantage to be a threat to agricultural production (Owen 2005). Further, soybean is almost exclusively self pollinated (Owen 2005).
Cotton has been grown in Australia for over 200 years. Approval for the introduction of GM cotton in Australia required a risk assessment of the environmental impact, including potential to become a weed and any effects on Australian native cotton species. None of the Australian native species are considered weeds and their distribution is relatively sparse and contact with GM cotton is rare (Brown et al. 1997). Further, cotton is predominantlyself pollinating and any hybrids that have been artificially formed between cultivated cotton and the native cotton species are completely sterile and therefore the risks of genes from GM cotton escaping into the endemic speciesis considered negligible (Constable et al. 1998; Fitt 2003). No natural hybrids between GM cotton and the native species have been reported (Fitt 2003).
Herbicide-tolerant GM and non-GM canola are considered to have a range of wild relatives that also grow in their areas of cultivation and hence pose a potential threat to agricultural production (reviewed by Beckie et al. 2003; Beckie et al. 2006; Hall et al. 2005; Légère 2005). In Canada, there are four wild species that have the ability to cross with canola: bird's rape otherwise known as field mustard; wild radish; wild mustard; and dog mustard (reviewed by Metz et al. 1997; Rieger et al. 1999). Concerns raised by opponents of GM crops relate to a possible fitness advantage that might be conveyed onto theseweed relatives, which also have the ability to further outcross and disseminate the trait(s) and hence cause inappropriate environmental impact (Rieger et al. 2002; Snow 2002).
The frequency of gene flow from GM canola to wild relatives was examined in both glasshouse and field experiments and indicates that the probability of gene flow from GM canola to a weedy relative is very low (for example, Warwick et al. 2003). Enrichment of any herbicide-tolerant weeds that might be generated is also considered unlikely to occur and persist in the environment unless a herbicide selection pressure was applied (Crawley et al. 2001). Therefore, any herbicide resistance gene that moves into non-agricultural environments would not increase the fitness of a plant in a natural environment where it is unlikely to be treated with herbicide (Crawley et al. 2001).
Rieger et al. (2002) quantified, at a landscape level, gene flow from herbicideresistant non-GM canola to nearby crops not containing herbicide resistance genes. This study provided the first evaluation of gene flow on commercial scale canola fields and sampled a range of environments over one-third of Australia. The authors found that pollen flow occurred at a very low frequency with less than 1 per cent observed between adjacent commercial canola fields. Herbicide resistance was not detected in canola fields more than 3 kilometres from a source field.
The impact of 10 years of commercialisation of herbicide tolerant crops in Canada was recently reviewed (Beckie et al. 2006). The authors did not observe marked changes in volunteer canola weed problems associated with herbicide-tolerant crops (GM and non-GM), except in reduced till systems when glyphosate was used alone to control canola volunteers. In this case, several weeds with inherently high glyphosate tolerance tended to associate with the continuous use of glyphosate (Harker et al. 2006). This 'evolved' weed resistance is not attributed to GM traits, but due to selection pressure associated with herbicide application per se. This is generally considered to be of greater environmental concern than resistance in related weed species caused through gene flow (Beckie 2006).
Both GM and non-GM herbicide resistant crops allow farmers to manage many difficult in-crop weeds in a cost effective manner (Brookes and Barfoot 2006a). However, management of herbicide tolerance amongst agriculturally important weeds remains an important challenge for modern agriculture (reviewed by Beckie 2006). Data to date suggest that concerns associated with herbicide resistance (as a result of the use of herbicide resistant GM and non-GM crops) are more related to crop management practices to control weeds than as a result of the herbicide resistance trait itself (Beckie 2006).
Impacts on diversity and natural habitats
The impact of herbicide resistant crops on weed diversity has also been studied in field crops of GM beet, maize and canola and revealed that weed diversity was not affected (Heard et al. 2003). In Canada, a reduction in weed diversity has not been demonstrated (Beckie et al. 2006).
A major concern raised by opponents of GM crops is the potential for traits to invade and naturalise in natural habitats and affect landscape diversity. Some concern was raised about reports that GM material from GM corn was found in Mexican land races leading Mexican authorities to cease growing commercial GM corn (reviewed by Hodgson 2002). Despite concerns, no long term introgression of GM traits into wild populations leading to the extinction of any wild species has been reported. This may be, in part, explained by the observation that the transition of vegetation from agricultural landscapes into natural habitats is usually gradual and with distance comes less risk (Sanvido et al. 2006). Furthermore, it is suggested that herbicide-tolerant GM crops are no more likely to be invasive in agricultural fields or natural habitats than non- GM plants and that a selective advantage for GM crop volunteers is not likely outside agricultural production areas (Crawley et al. 2001).
A6.5 Ecological benefits of GM crops
A number of studies have considered the ecological benefits associated with the introduction of GM crops (for example, Ammann 2005; Beckie et al. 2006; Benbrook 2003; Brookes and Barfoot 2006 a and b; Constable et al. 1998; Dale et al. 2002; Fernandez-Cornejo and McBride 2002; Fitt et al. 2004; Knox et al. 2006; Marvier et al. 2007; Sanvido et al. 2006; Vadakattu and Watson 2004). These reports indicate that the numbers and amount of pesticide applications have significantly reduced in every country where Bt cotton has been adopted and more broadly a significant reduction in environmental impact associated with GM crop adoption. Direct environmental benefits have been reported such as reduced effects to non-target organisms (Head et al. 2005; Marvier et al. 2007; Torres and Ruberson 2005) and reduced leachate into ground water (FAO 2004). Flow-on benefits from reduced pesticide applications includes a reduced risk to human health and safety through chemical exposure and application (Hossain et al. 2004; Warnemuende et al. 2007).
A6.5.1 Global reduction in pesticide use through GM crops
GM crops have contributed to a significant reduction in the global environmental impact of production agriculture (Brookes and Barfoot 2006b). Since the introduction of GM crops, the use of pesticides in agriculture has reduced by 224 million kg of active ingredient and the overall environmental impact has reduced by 15.3 per cent.
The volume of herbicide use in GM soybeans has decreased by 51 million kg with the overall environmental impact decreased by 20 per cent. It should be noted that in some countries, such as in Argentina, the adoption of GM soybeans has coincided with increases in the volume of herbicides used relative to historic levels (Trigo and Capp 2006). This largely reflects the adoption of reduced tillage production systems with their inherent environmental benefits (Brookes and Barfoot 2006b; Trigo and Capp 2006).
Major environmental gains have also been derived from the adoption of GM insect-resistant cotton with a 95.5 million kg reduction of insecticide and a 24.3 per cent reduction in environmental impact. Additional environmental gains have also come from GM corn and GM canola (Brookes and Barfoot 2006b). In corn, pesticide use has decreased by 43 million kg and the environmental impact decreased due to a combination of reduced insecticide use (4.6 per cent) and a switch to more environmentally-benign herbicides (4 per cent). Similarly, in the canola sector, farmers have reduced herbicide use by 6.3 million kg (an 11 per cent reduction) and the environmental impact has fallen by 23 per cent, primarily due to a switch to more environmentallybenign herbicides (Brookes and Barfoot 2006b).
Benbrook (2003) stated that overall pesticide use in the United States actually increased by 50.6 million pounds (approximately 23 000 tonnes) through the adoption of GM crops from 1996–2003. It was suggested that the average amount of herbicides applied per acre planted to herbicide resistant varieties increased over the period compared with the first few years of adoption, whilst Bt varieties reduced pesticide use by 19.6 million pounds (8900 tonnes). These data have been widely criticised as the amount or toxicity of the active ingredient has not been taken into account. When these are considered, pesticide use rates in the United States on corn, soybeans and cotton has declined by 2.5 million pounds (1100 tonnes), despite an increase in the totalamount of herbicides applied to GM soybeans (Fernandez-Cornejo and Caswell 2006). In addition, the switch from the pre-emergence herbicide metolachlor to glyphosate in soybean production has had a huge benefit to the environment that cannot be measured in terms of the amount of active ingredient alone. Instead, the benefits are measured in a decline in groundwater contaminants and a reduction in risks to human health and safety (Cerdeira and Duke 2007). Similar benefits might be expected in Australia through a reduction in the use of triazine-based herbicides. It has also been reported that the switch to glyphosate has helped foster soil conservation practices such as the use of reduced tillage agriculture, leading to additional benefits through reduced soil erosion, decreased on-farm fuel consumption, and improved wildlife habitat (reviewed by Duke 2005; Fawcett and Towery 2002).
A6.5.2 Benefits of GM cotton in Australia
The first, single gene, Bt cotton, marketed as INGARD®, was made commercially available in Australia in 1996. In 2002, a two-gene Bt cotton was introduced in Bollgard®II varieties followed by a rapid phase out of INGARD® varieties after the 2003-04 season. The two-gene varieties were seen to provide greater efficacy and resilience against the risk of insects developing resistance to the effects of the single Bt gene varieties (Bates et al. 2005; Roush 1998).
The environmental benefits of reduced insecticide use associated with GM cotton adoption in Australia have been assessed (Knox et al. 2006). An environmental impact quotient (EIQ) was used to assign a potential hazard value to the agricultural pesticides used in cotton production. The EIQ assesses farm worker, consumer and environmental risk and then combines these to provide an EIQ value for a particular pesticide (Kovach et al. 1992). The environmental impact of a pesticide can then be determined by multiplying the EIQ by the amount (kg/ha) of active ingredient. Results from an evaluation of the environmental impact associated with GM and non-GM cotton indicated that the shift in pesticide use has led to more than a 64 per cent reduction compared with non-GM cotton. The average environmental impact for 2002-03 and 2003-04 for non-GM cotton was 135 kg active ingredient per hectare compared with 28 kg active ingredient per hectare for the two-gene Bt cotton variety (Knox et al. 2006).
In 2006, the Gene Technology Regulator issued licences (DIR 059/2005; DIR 062/2005; DIR 066/2005) for the phased introduction of GM cotton containing resistance to the herbicides glyphosate and glufosinate ammonium in addition to the two Bt genes. The associated changes in crop management could result in a further reduction in environmental impact compared to non-GM cotton.
Cognisant of the need for rigorous nutritional and safety assessment, regulatory agencies worldwide have established case-by-case risk assessments to evaluate potential impacts of GM food crops on human and animal health and the environment. Concepts and principles for safety have been developed by international organisations such as the World Health Organization (WHO), the Food and Agriculture Organization (FAO) of the United Nations, the Organisation for Economic Co-operation and Development (OECD) and the Codex Alimentarius Commission. In Australia, the Office of the Gene Technology Regulator (OGTR) has developed a risk analysis framework that describes the principles of risk analysis used to assess GM applications with respect to protection of the safety of human health and the environment (see appendix 4). Food Standards Australia New Zealand (FSANZ) examines any differences between an existing food and a new GM product (that is, an assessment of 'substantial equivalence'). FSANZ assessments investigate toxicity, any tendency to provoke an allergic reaction, the stability of the inserted gene, whether there is any nutritional deficit or change, and unintended effects of the inserted genes, such as changes in known naturally occurring toxins. A GM food will only be approved for sale in Australia and New Zealand if it is as safe and nutritious as its conventional counterparts.
The introduction of GM crops has provided a number of benefits to those farmers and countries that have adopted them. These include new agronomic traits that have overcome several production constraints leading to improvements in productivity and profitability. Further benefits to the environment from changes in crop management are also being realised. However, the adoption of GM crops has also raised some new challenges for agricultural management: stewardship and ongoing rigorous case-by-case assessment of the safety for human and animal health and the environment are fundamental in ensuring the sustainability of agriculture and the environment into the future.
 Bacillus thuringiensis, commonly known as Bt, is a soil bacterium that produces proteins that are toxic to specific Orders of insects. When an insect ingests the Bt protein, the function of their digestive system is disrupted, producing slow growth and ultimately death. Bt is very selective and effective against important agronomic pests such as European corn borers and cotton bollworms (Lepidoptera), Colorado potato beetles (Coleoptera), and certain flies and mosquitoes (Diptera).