Implementation and Impact of Transgenic Bt Cottons in Australia

Transcription

1 Implementation and Impact of Transgenic Bt Cottons in Australia Gary P. Fitt* Australian Cotton Cooperative Research Centre and CSIRO Entomology, PO Box 59, Narrabri,NSW, Australia, 2390 Pp in Cotton Production for the New Millennium. Proceedings of the third World Cotton Research Conference, 9-13 March, 2003, Cape Town, South Africa, 1778 pages. Agricultural Research Council - Institute for Industrial Crops, Pretoria, South Africa. * Current Address: CSIRO Entomology, LongPocket Laboratories, 120 Meiers Rd, Indooroopilly, Qld, 4068

2 Implementation and Impact of Transgenic Bt Cottons in Australia Gary P. Fitt Abstract Genetically engineered cottons expressing the delta-endotoxin genes from Bacillus thuringiensis subsp. kurstaki (Bt) offer a significant step forward in cotton pest management. Bt cotton varieties have been now been commercialised in nine countries to combat various lepidopteran pests, principally the Heliothines. These Bt cottons offer great potential to dramatically reduce pesticide use for control of the major Lepidopteran pests and offer a valuable tool for sustainable IPM systems. In Australia, Bt cottons (tradename INGARD ) expressing the CryIAc endotoxin, were commercialised in 1996/97 and have gradually increased in area under an industry agreed deployment strategy which limits the use of the first generation of Bt cottons to 30% of the cotton area. All Bt varieties are grown under a comprehensive management strategy designed to minimise the risk of resistance evolving in Helicoverpa armigera, the main target pest. The mandatory use of refuges, crop residue destruction and planting windows are all components of that strategy. In commercial use Bt cottons have reduced pesticide needs for Helicoverpa armigera and H. punctigera by an average of 56%, providing major environmental benefits. Yield potential and fibre quality have been maintained and in some cases significant yield increases have been documented. In economic terms, the first 4 years of adoption of Bt cottons produced average net returns to growers similar to conventional cotton, due to the licence fee and the additional costs associated with the management strategy. During that time there was considerable variability in performance of Bt crops. In the last two years Bt cottons have provided significant economic advantage, due to better performing varieties, greater management experience with the technology and relatively low pest abundance. Despite initial difficulties growers have rapidly adopted the technology and are likely to benefit in the long term through the clear contribution of Bt cottons to environmental sustainability. With Bt cottons restricted to only a proportion of the cotton area growers have selectively used them in environmentally sensitive areas near watercourses or townships. Perhaps more importantly Bt cottons have shown great potential as a component of IPM systems where a combination of soft approaches yields considerable benefits. That single gene Bt cotton has not been a magic bullet is somewhat fortunate in demonstrating that pest tolerant transgenic cottons need to be introduced as part of an IPM system and supported by rigorous and well-researched resistance management strategies. The recent (2003/04 season) introduction of varieties expressing a second protein (Cry2Ab) will see enhanced efficacy but not reduce the need for careful management and an IPM approach to ensure sustained benefit from this application of biotechnology.

3 Introduction Cotton is a significant primary industry for Australia with some 500,000 ha grown, producing 3.0 to 3.4 million bales annually. Over 90% of production is exported as raw lint making cotton the third largest crop export and Australia the fourth largest player in the world export market. Yields and fibre quality among the worlds highest (Hearn and Fitt 1992). A number of insect pests attack the crop (Fitt 1994, 2000). Historically pest management on conventional cotton has relied largely on synthetic insecticides, although some biological approaches are also utilised (Fitt 1994). This dependence on pesticide intervention brings with it several environmental and economic liabilities, including the development of insecticide resistance in key pests and secondary pests, reductions in beneficial insect populations with associated secondary pest outbreaks, potential for environmental contamination and cost - pest management now accounts for 35-40% of the variable costs of production ($400-$1000/ha) (Fitt 2000). To reduce these problems of pesticide dependence the Australian industry is increasingly committed to Integrated Pest Management. Genetically engineered cottons expressing the delta-endotoxin genes from Bacillus thuringiensis subsp. kurstaki (Bt) offer great potential to dramatically reduce pesticide dependence for control of the major Lepidopteran pests and consequently offer real opportunities as a component of sustainable and environmentally acceptable IPM systems. Bt cotton varieties have now been commercialised in nine countries worldwide: USA, Australia, China, India, Indonesia, Argentina, Mexico, South Africa and Colombia (James 2002). Australian Bt cottons (tradename INGARD ) utilise the CryIAc gene from Monsanto and deployed in most commercial cotton varieties released by Cotton Seed Distributors and Deltapine. INGARD varieties were first registered in 1996 when a limited commercial release was approved. Since that time INGARD varieties have significantly reduced pesticide needs for the major Lepidopteran pests, Helicoverpa armigera and H. punctigera. In this paper I discuss the introduction of INGARD technology in Australia and the impacts this has had on producers, researchers, and the industry as a whole. Pests of Australian cotton Primary pests are the larvae of two noctuid moths, Helicoverpa punctigera Wallengren and H. armigera Hubner, with a range of secondary sucking pests and mites (Pyke and Brown 1996). H. punctigera remains susceptible to the main insecticide classes, while H. armigera has evolved high levels of resistance to several classes including to the pyrethroids, endosulfan, carbamates and has incipient resistance to the organophosphates (Forrester et al. 1993). While both Helicoverpa spp. have experienced only slight exposure to Bt sprays on cotton they are both naturally much more tolerant of Cry IA toxins than is Heliothis virescens, the main target for BOLLGARD cotton in the USA (N.

4 Forrester pers. comm.). This difference introduces additional difficulties in achieving a high and consistent level of efficacy of INGARD cotton varieties. INGARD Cotton pre-commercial release Table 1 summarises the major field research and development stages for INGARD Cotton leading to limited commercial release in 1996/97 and then phased introduction thereafter. Between 1992/93 and 1995/96, the field assessments were controlled and monitored by the Genetic Manipulation Advisory Committee (GMAC), a Federal government advisory body which oversaw all contained research and field releases of transgenic organisms and provided advice to the National Registration Authority as part of the commercial approval of INGARD cotton in 1996/97. INGARD registration was made conditional upon the development of an industry approved resistance management strategy. This was accommodated through a cotton industry committee, the Transgenic and Insecticide Management Strategy Committee (TIMS), with a membership consisting of cotton growers, scientists, cotton consultants and chemical industry representatives. The TIMS Committee is a subcommittee of the Australian Cotton Growers Research Association (ACGRA) which represents the interests of growers in research and technical issues. TIMS is responsible for developing, promoting and reviewing resistance management strategies in cotton for both transgenic plants and conventional pesticides. From 2001, the legislated Office of the Gene Technology Regulator (OGTR) assumed responsibility for the oversight of issues with GM crops in areas of human health and environmental safety. Insert Table 1 about here Despite the potential benefits of Bt cotton technology and the demonstrated safety of Bt in conventional sprays, a number of potential environmental and ecological impacts required research prior to commercial release. These were addressed through a series of field and laboratory studies (Fitt et al 1994, Llewellyn and Fitt 1996, Brubaker et al 1999, Fitt and Wilson 2002, Table 2), which in some cases are ongoing. Research continues on potential impacts of Bt cotton residues on soil microbes and potential weediness of Bt varieties in tropical environments of northern Australia (Eastick 2002). Insert Table 2 about here Commercial Performance of INGARD cottons Pyke and Fitt (1998) provide a summary of the performance of INGARD cotton varieties over the first two years of commercial use and the interaction of the industry with Monsanto. Comprehensive information on pesticide use, yields and returns to growers is provided in reports produced by the Cotton Research and Development Corporation (Long et al 1997, Clark et al 1998, Pyke 1999, Kwint and Pyke 2000, Doyle et al 2002, 2003). Table 3 summarises the key points from these comparative analyses around the issues of yield, pesticide use, pest control costs and economic returns.

5 [Insert Table 3 about here] Prior to release and in the first year of commercial use there were a number of unrealistic expectations of INGARD technology, which were reinforced by the initial licence fee of $245 per hectare. In the first year of field research in Australia, 1992/93, it was demonstrated that Bt transgenic cotton plants were very effective in controlling Helicoverpa larvae for much of the growing season, but did not provide season long control. There was evidence of progressive decline in efficacy as plants matured and senesced (Fitt et al, 1994). This observation set the scene for much research which followed, seeking to quantify the seasonal changes in efficacy of INGARD varieties and the environmental/agronomic factors which may influence efficacy. Subsequent experience in Australia has confirmed that efficacy of varieties expressing CryIAc is not consistent through the growing season and can be highly variable (Fitt et al 1994, Fitt et al 1998). Efficacy against Helicoverpa spp. typically declines through the boll maturation period, to the point where survival of larvae is little different to that in non-transgenic cotton (Fitt et al. 1994, Fitt et al 1998), although growth rates of survivors on the INGARD crops are still dramatically reduced (Fitt unpublished). The decline in efficacy necessitates supplementary Helicoverpa control on INGARD crops, particularly in the last third of the growing season. The first commercial year also highlighted the variability in performance. Given its importance in managing pests on INGARD cotton the causes of variable efficacy has been the subject of a considerable research effort. In the 1997/98 season Monsanto retained a licence fee of $245 per ha but with rebates totaling $35 per ha for growers who complied with various components of the resistance management strategy. A form of Value Guarantee was reintroduced following expressions of dissatisfaction with INGARD performance from some growers, particularly in areas where Helicoverpa numbers were extreme. In subsequent years the INGARD licence fee dropped to $185/ha with a $30 rebate for full compliance with the resistance management strategy, giving a net fee of $155/ha which has remained largely unchanged since. There is however, huge variation in returns with some growers in each year achieving gains of $1000/ ha while others record a net loss of $1000/ha (Figure 2). The most consistent winner from INGARD technology to date has been the environment, with reduced pesticide loads, while the cotton industry has gained long term sustainability through the progressive adoption of more integrated pest management approaches using INGARD cotton as a foundation. Impact of INGARD cotton on Pesticide Use

6 Figure 1 summarises the average spray numbers applied to INGARD and conventional crops for a range of cotton pests. For Helicoverpa, there have been consistent reductions in the number of sprays applied to INGARD crops compared to conventional ones, averaging 56% and ranging from 43% (1998/99) to 80% (2001/02). For the minor pest groups there has been no significant change in pesticide applications after 6 years. In 1998/99 it is estimated that the reduction in spraying across 125,000 ha of INGARD crops resulted in 1.75 million litres less pesticide entering the environment. Despite the variable performance, the average 56% reduction in pesticide applications for Helicoverpa represents a spectacular impact for an IPM product. [insert Figure 1about here] On average the greatest reduction in sprays has been during the squaring (flowerbud) and flowering stages of crop development (50-67%) whereas reductions during boll filling and opening have been more modest (20-35%) (Table 4). Figure 2 summarises changes in usage of particular chemical groups. The most spectacular reductions were in early season use of endosulfan and in the carbamates and synthetic pyrethroids. In each case this reduction has major environmental benefits. [insert Table 4 about here] [insert Figure 2 about here] Economic Impacts of INGARD cotton With significant reductions in pesticide applications, INGARD cotton could be expected to provide greater returns to growers. As yet there has been no comprehensive economic assessment of the value of INGARD technology in the Australian environment, in contrast to those in the USA (eg Falck-Zepeda et al 2000), which show significant benefit accruing to growers. In Australia much of the economic benefit in the first 4 years was subsumed by the technology licence fee, with the result that net economic return was similar to conventional cotton (Table 3). In the last two years net economic returns from INGARD varieties have been considerably higher (Table 3) at over $300/ha. This reflects the progressive improvement in varietal performance, growing experience of growers and consultants in managing INGARD cottons and the lighter pest pressure experienced in those years. Nonetheless intangibles due to reduced environmental impact and enhanced sustainability have not yet been valued. Over the six years of commercial use there have been changes in licence fee, area of INGARD grown, range of varieties available and pest pressure, all of which complicate between year comparisons. In the first year (96/97), INGARD cotton was $43 per ha more expensive to produce than conventional cotton (ie. the combined cost of the licence fee with the cost of sprays applied). In addition average yields for INGARD crops were 0.53 bales per ha lower yielding than the

7 conventional comparisons in that year. However, this was largely due to the limited availability of seed for all varieties which resulted in a significant number of growers using INGARD varieties unsuited to their region. In subsequent seasons varietal choice has not been a limitation. In general average yields of INGARD have been comparable to or higher than conventional cottons, although in the 1998/99 season yields of INGARD varieties were markedly lower in one region (Gwydir). There were no significant differences in crop maturity (time from planting to harvest) between INGARD and conventional varieties in any year. In the first 3 years of INGARD use average costs per insecticide spray were higher for INGARD crops, but little or no difference subsequently. This effect resulted from growers and consultants applying recommended thresholds for INGARD crops which required that the threshold be met on two consecutive checks of an INGARD crop, whereas control would be triggered after one over threshold check on a conventional crop. In INGARD crops which were performing poorly the threshold often allowed larvae to reach larger sizes before action was taken. Consequently more expensive, hard chemicals were sometimes required to regain control. With more experience growers and consultants have adjusted their expectations of INGARD performance from mid season. Evenso some 80-90% of consultants still apply the recommended thresholds over two consecutive checks of INGARD crops. Table 4 also shows net insect control costs were more expensive on INGARD crops in the first two years, largely due to the elevated technology licence fee which applied in those years. Net economic benefit in subsequent years was slightly positive due to the reduced licence fee (to $155/ha net) and insect control costs being $70-$90 lower on INGARD crops. As the data shows however, there was considerable variation among crops in economic benefit, particularly in the last two years when benefits ranged from +$1000/ha to -$1000/ha (Figure 3). In part this is due to: differences in pest pressure in different regions, to greater numbers of growers using INGARD varieties resulting in greater divergence in experience of managing the crop, and in part to the variation in efficacy among INGARD crops. [insert Figure 3 about here] Despite the variability of economic return and additional requirements of INGARD cotton growers, it is noteworthy that all the licensed area of INGARD cotton available each year was sold. Clearly growers attribute value to the environmental benefits associated with INGARD cotton. Indeed some 80-90% of growers have consistently identified their prime reason for growing INGARD cotton as protection of the environment. Since the 1999/2000 season, by which time expectations of INGARD cotton were more realistic, over 70% of consultants and growers have indicated they believe INGARD cotton provided even or better value for money and some 80% of surveyed growers were satisfied with INGARD performance.

8 Causes and implications of variable efficacy Prior to field trials, it had been expected that expression of the Bt protein would be consistent throughout growth of the cotton crop. The Cry1Ac gene is driven by a CMV35S promoter which gives constitutive expression in all tissues of the plant, although there are significant differences between plant structures in the level of Bt protein production (Benedict et al. 1996, Fitt and mares unpublished). However, from the first year of small-scale field trials, it became evident that efficacy of leaves and reproductive tissues declined during plant growth (Fitt et al. 1994) and that some larvae were able to survive beyond first instar late in the season Fitt et al (1998), Daly and Fitt (1998) and Holt (1998) summarise information on the seasonal patterns of efficacy against H. armigera. Considerable research using well-validated bioassays techniques and ELISA (Holt et al 2002) is quantifying changes in the concentration of Bt toxins over time and the factors which influence that. Coupled with the variability of efficacy between crops, the consistent seasonal decline in efficacy has several implications. One relates to the uncertainty generated for crop managers and the confidence they may have in using thresholds for Bt cotton. Current thresholds are 2 small larvae/metre of row for two consecutive check dates (or 1 medium larva/metre on the first check). Requiring that the threshold be exceeded on two consecutive checks may result in pesticides being applied to larger larvae which are more difficult to control. While surviving larvae on Bt plants grow more slowly than larvae on conventional cotton (and many of them die eventually even if insecticide sprays are not applied), they still cause damage. Some crops have suffered damage because consultants complied with thresholds and left small larvae untreated; subsequent insecticide treatments failed to control larger larvae. An unfortunate consequence is that consultants may become more conservative in their use of thresholds (some even reverted to egg thresholds with Bt cotton), thereby eroding the full realisation of IPM value of INGARD cotton. A second significant consequence of changing field efficacy is the added selective pressure this may apply on Bt resistance. The resistance management strategy in place for Bt cotton (Roush et al 1998) relies on the use of refuges but these work best when the plants are highly efficacious and have the capacity to kill a high proportion of heterozygous resistant individuals. INGARD cotton varieties expressing a single Cry1Ac gene clearly do not express a high dose and heterozygote mortality is unlikely to be high except perhaps when plants are quite young (up to squaring phase). A significant period of selective mortality is likely during the season. Given this, the other components of the resistance management strategy are even more crucial. It is for this reason that use of INGARD cotton has been restricted by a cap on area in the first years of commercial use (discussed more fully later). This serves to magnify the size of the total refuge which includes the refuge crops grown specifically with the Bt cotton, the area of conventional cotton (at least 70% of the total) and all other Helicoverpa susceptible crops grown in eastern Australia. This represents a huge refuge.

9 It is worth noting that the species targeted by INGARD cotton in Australia, H. armigera and H. punctigera, are much more tolerant (10 fold more) of the Cry1Ac protein than the main target of BOLLGARD cotton (same Cry1Ac protein) in the USA industry, Heliothis virescens. Resistance Risk and Management Requirements The major challenge to sustainable use of transgenic cottons is the risk that target pests may evolve resistance to the CryIAc protein. Resistance to conventional Bt sprays has evolved in field populations of Plutella xylostella (Tabashnik, 1994a,b) and the possibility of resistance is a real concern, particularly for H. armigera which has consistently developed resistance to synthetic pesticides (Forrester et al., 1993, Fitt, 1989, 1994). For this reason a pre-emptive resistance management strategy was developed and implemented to accompany the commercial release of transgenic varieties. The strategy adopted in Australia is targeted at H. armigera and based on the use of refugia to maintain susceptible individuals in the population (Roush, 1996, 1997; Gould, 1994). This strategy seeks to take advantage of the polyphagy and mobility of Helicoverpa spp. to achieve resistance management by utilising gene flow to counter selection in transgenic crops. Key elements of the strategy are as follows: effective refuges on each farm growing INGARD cotton defined planting window for INGARD cotton to avoid late planted crops that may be exposed to abundant H. armigera late in the growing season mandatory cultivation of INGARD crops to destroy most overwintering pupae of H. armigera defined spray thresholds for Helicoverpa to ensure any survivors in the crops are controlled monitoring of Bt resistance levels in field populations All elements of the INGARD management plan are included in a printed strategy provided to all cotton growers and available on the web ( The plan stipulates planting times and areas for refuges, necessary distances of refuge from INGARD crop, and management requirements for refuges. Refuge options have been defined on the basis of ongoing research which seeks to quantify the value of different options in generating moths (eg. Fitt and Tann 1996) and ranks potential refuges in relation to unsprayed conventional cotton, regarded as the control refuge. Options are expressed as the number of hectares required for every 100ha of INGARD cotton. Current refuge options include: 10 ha of unsprayed conventional cotton 100 ha of sprayed conventional cotton 15 ha of unsprayed sorghum 20 ha of unsprayed corn 5 ha of unsprayed pigeon pea

10 The major criteria defining effective refugia are that they not be treated with Bt sprays, and that they generate enough susceptible adult Helicoverpa to ensure that matings between two resistant survivors from a transgenic crop are extremely unlikely events. Refuges are required to be in close proximity to the transgenic crops (within 2 km) to maximise the chances of random mating among sub-populations (Dillon et al 1998). At present the conventional sprayed cotton refuge option remains most popular, although an increasing proportion of growers are now using unsprayed non-cotton refuges (40% in 01/02). Unsprayed refuges will be required for the future extensive deployment of BOLLGARD II TM cottons expressing both Cry IAc and Cry 2Ab proteins. An additional element of the conservative Australian deployment of Bt cotton is the imposition of a phased introduction of INGARD varieties and a cap on the area at 30% of the total cotton area. In their first year INGARD varieties were grown on 30,000 ha representing about 8% of the total cotton area in that year. After the first season the TIMS committee introduced a cap in response to an industry desire for a phased introduction (increasing in 5% increments each year). A cap of 30% was suggested by growers and supported by simulation modeling as a means of further minimising the risk of resistance to the CryIAc protein before the future release of two gene combinations (CryIAc/ Cry2Ab). The 30% cap, first applied on a regional basis, now applies at farm level and was first reached in the 2000/2001 season with 165,000ha registered for use. BOLLGARD II TM varieties have now been approved for commercial use and occupy about 5,000 ha in 2002/03. The two gene varieties provide much better efficacy and hence even greater reduction in pesticide requirement, but their main purpose is to provide much greater resilience against the risk of resistance (Roush 1998). A rapid transition to BOLLGARD II TM is planned, with INGARD varieties likely to be withdrawn after the 2003/04 season. Compliance and Auditing All aspects of the Insect Management Plan for INGARD cotton are embodied in the label and are part of a single use contract growers must sign with Monsanto in order to purchase seed. Components of the resistance management plan are thus legally binding on the grower. To support this the contract and label also stipulate that each farm growing INGARD cotton be audited three times each year, with audit visits in December, February and August to check on compliance with refuges, pesticide use and compulsory plowdown. Since 1999/2000 the technology licence fee for INGARD cotton has also included a rebate for growers who meet the requirements of all three audits. Cotton varieties with INGARD were registered by the National Registration Authority for an initial period of 5 years. The registration includes an annual review with a requirement that all audit information be provided. The annual registration review also allows the NRA to approve a registered area for the next year based on advice from the TIMS committee on the % area to be grown (up to 30%

11 of the predicted total cotton area). Monsanto is then responsible for an allocation process which ensures that no more than 30% of a region is INGARD cotton. The Office of the Gene Technology regulator is currently considering a renewal of the 5-year registration. Transgenic Bt cotton as a foundation for IPM INGARD and BOLLGARD II TM cotton varieties are not perceived as magic bullets for pest control in Australia. Instead they are viewed broadly as an opportunity to address environmental concerns about cotton production and more specifically as a foundation to build IPM systems which incorporate a broad range of biological and cultural tactics (Fitt 2000). Research has shown little effect of Bt cottons on non-target species, including non-lepidopterous pests, beneficial insects, and other canopy dwelling and soil dwelling species (Fitt et al. 1994; Fitt and Wilson 2000; Fitt and Wilson 2002). Survival of beneficials is markedly higher than in conventional sprayed cotton, and they provide control for some secondary pests, particularly those which are induced pests in sprayed cotton (eg. mites and aphids). Reduced use of disruptive pesticides will allow greater focus on the management and manipulation of beneficial species, using nursery crops and food sprays (Fitt 2000), or other means of conservation and augmentation. A range of other tactics can also be deployed. These would include selective pesticides, behaviour modifying compounds, conventional HPR characters. These themes are discussed more fully in others papers in this proceedings. IPM systems for future cotton production will, of necessity, be more complex than the pesticide based systems currently in place, and will require greater effort on the part of crop managers (Wilson et al 1998, Wilson et al 2003). In summary the key benefit for the development of IPM offered by INGARD technology is a reduced need for season long control of a primary pest with broad-spectrum insecticides. Conclusions INGARD cotton, as the first commercial introduction of biotechnology to the Australian cotton industry, has shown the potential of transgenic pest tolerant crops to significantly reduce pesticide use, providing major environmental benefits. After an initial period of negligible economic benefits, INGARD varieties are now returning significant economic benefit. At this time it is unknown what the licence fee for BOLLGARD II varieties may be. Despite this growers are likely to benefit in the long term through the clear contribution of Bt cottons to sustainability. Until now INGARD cotton has occupied only a restricted proportion of the Australian cotton area. With BOLLGARD II TM varieties the 30% cap is likely to be removed and substantially larger areas are likely. That INGARD cotton has not been a magic bullet is fortunate in demonstrating that pest tolerant transgenic cottons will need to be introduced as part of an IPM system that incorporates a range of tactics. The deployment of

12 BOLLGARD II cotton and a range of herbicide tolerant cottons will see transgenic varieties become an important cornerstone of sustainable cotton production systems for Australia.. References Brown, A. H. D., C. L. Brubaker, and M. J. Kilby Assessing the risk of cotton transgene escape into wild Australian Gossypium species. In G. D. McLean, P. M. Waterhouse, G. Evans, and M. J. Gibbs [eds] The Commercialisation of Transgenic Crops: Risk, Benefit and Trade Considerations [Proceedings of a workshop held in Canberra Australia March 1997], pp Cooperative Research Centre for Plant Science and Bureau of Resource Sciences, Canberra, Australia. Brubaker, C. L., A. H. D. Brown, J. McD. Stewart, M. J. Kilby, and J. P. Grace Production of fertile hybrid germplasm with diploid Australian Gossypium species for cotton improvement. Euphytica 108: Clark D., Long, T. and Pyke, B The performance of INGARD cotton in Australia in the 1997/98 season. Cotton R&D Corporation Occasional Paper. 51pp. Daly, J.C. and Fitt, G.P Efficacy Of Bt Cotton Plants In Australia - What Is Going On? Pp in New Frontiers in Cotton Research - Proceedings World Cotton Research Conference - 2; Athens, Sept ; Editor: F. Gillham; Publishers: P. Petridis, Thessaloniki Dillon, M.L., Fitt, G.P. and M.P. Zalucki How should refugia be placed upon the landscape? A modelling study considering pest movement and behaviour. pp In M.P. Zalucki, RAI Drew, GG White (eds). Pest Management - Future Challenges. Proceedings 6th Australasian Applied Entomology Conference Vol. 1, Brisbane. University of Qld Press, Brisbane. Doyle, B., Reeve, I., Barclay, E. (2002). The performance of INGARD cotton in Australia in the 2000/01 season. Institute for Rural Futures, University of New England. 50pp. Doyle, B., Reeve, I., Bock, K. (2003). The performance of INGARD cotton in Australia during the 2001/02 season. Institute for Rural Futures, University of New England. 82pp. Eastick, R. (2002) Evaluation of the Potential Weediness of Transgenic Cotton in Northern Australia. Falck-Zepeda J.B., Traxler G., Nelson R.G. (2000) Surplus distribution from the introduction of a biotechnology innovation. American Journal of Agricultural Economics 82:

13 Fitt, G.P The ecology of Heliothis species in relation to agroecosystems. Annual Review of Entomology 34: Fitt, G.P Cotton pest management: Part 3. An Australian perspective. Annual Review of Entomology 39: Fitt, G.P An Australian approach to IPM in cotton: integrating new technologies to minimise insecticide dependence. Crop Protection 19: Fitt, G.P. and C. Tann Quantifying the value of refuges for resistance management of transgenic Bt-cotton. In Proceedings of the Eighth Australian Cotton Conference, Broadbeach, August, pp Fitt, G.P., Daly, J.C., Mares, C.L. and K. Olsen Changing Efficacy of Transgenic Bt Cotton - Patterns and Consequences. pp In M.P. Zalucki, RAI Drew, GG White (eds). Pest Management - Future Challenges. Proceedings 6th Australasian Applied Entomology Conference Vol. 1, Brisbane. University of Qld Press, Brisbane. Fitt, G.P., Mares, C.L. and D.J. Llewellyn Field evaluation and potential impact of transgenic cottons (Gossypium hirsutum) in Australia. Biocontrol Science and Technology 4: Fitt, G.P. and Wilson, L.J Genetic Engineering in IPM: Bt cotton. pp In: Kennedy, G.G. and Sutton, T.B. (eds). Emerging Technologies in Integrated Pest Management: Concepts, Research and Implementation. APS Press, St. Paul. Fitt, G.P. and L.J.Wilson Non-Target Effects of Bt-Cotton: A Case Study From Australia. In Biotechnology of Bacillus thuringiensis and Its Environmental Impact: Proceedings of the 4 th Pacific Rim Conference, Akhurst, R.J., Beard, C.E., Hughes, P.A. (eds). CSIRO, Canberra. Pp Forrester, N.W., Cahill, M., Bird, L.J. and J.K. Layland Management of pyrethroid and endosulfan resistance in Helicoverpa armigera (Lepidoptera: Noctuidae) in Australia. Bulletin of Entomological Research Supplement No.1. Gould, F Potential and problems with multi-gene, high dose strategies for managing resistance to Bt toxins. Biocontrol Science and Technology 4: Hearn, A.B. and Fitt, G.P., Cotton cropping systems. In: Pearson C. (Editor), Field Crop Ecosystems of the World. Elsevier Press, Amsterdam, pp

14 Holt HE Season-long quantification of Bacillus thuringiensis insecticidal crystal protein in fieldgrown transgenic cotton. Pp In M.P. Zalucki, RAI Drew, GG White (eds). Pest Management - Future Challenges. Proceedings 6th Australasian Applied Entomology Conference Vol. 1, Brisbane. University of Qld Press, Brisbane. Holt H.E.; Mares, C.L., Akhurst R. (2002) Determination of the Cry protein content of Bt transgenic cotton a technical manual for laboratory use. CSIRO Entomology Technical Report No pp. James, C. (2002) Global Review of Commercialised Transgenic Crops: 2001 Feature: Bt cotton. ISAAA Briefs No pp. ISAAA: Ithaca, NY. Kwint, P. and Pyke, B. (2000) The performance of INGARD cotton in Australia in the 1999/2000 season. Cotton R&D Corporation Occasional Paper. 46pp. Llewellyn D.L.and Fitt, G.P Pollen dispersal from a field trial of transgenic cotton in the Namoi Valley, Australia. Molecular Breeding Long, A., Pyke, B. and Slack-Smith P The performance of INGARD cotton in Australia in the 1996/97 season. Cotton R&D Corporation Occasional Paper. 26pp. Pyke, B The performance of INGARD cotton in Australia in the 1998/99 season. Cotton R&D Corporation Occasional Paper. 46pp. Pyke, B. and G.P. Fitt Field performance of INGARD cotton - the first two years. pp In M.P. Zalucki, RAI Drew, GG White (eds). Pest Management - Future Challenges. Proceedings 6th Australasian Applied Entomology Conference Vol. 1, Brisbane. University of Qld Press, Brisbane. Roush, R. T Can we slow adaptation by pests to insect-resistant transgenic crops? pp In G. Persley (ed), Biotechnology and Integrated Pest Management, CAB International, London. Roush, R.T Managing resistance to transgenic crops. pp In N Carozzi and M Koziel (eds). Advances in insect control: The role of transgenic plants. Taylor and Francis, London. Roush, R.T Two toxin strategies for management of insecticidal transgenic crops: Can pyramiding succeed where pesticide mixtures have not? Philisophical Transactions of the Royal Society of London B. 353: Roush, R.T., Fitt, G.P., Forrester, N.W. and Daly, J.C. (1998) Resistance Management for Insecticidal Transgenic Crops: Theory and Practice. pp In M.P. Zalucki, RAI Drew, GG White (eds).

15 Pest Management - Future Challenges. Proceedings 6th Australasian Applied Entomology Conference Vol. 1, Brisbane. University of Qld Press, Brisbane. Tabashnik, B.E. 1994a. Delaying insect adaptation to transgenic crops: seed mixtures and refugia reconsidered. Proceedings Royal Society of London, Series B. 255: Tabashnik BE. 1994b. Evolution of resistance to Bacillus thuringiensis. Annual Review of Entomology 39: Wilson, L.J., Fitt, G.P. and R.K. Mensah. (1998). INGARD cotton - its role in cotton IPM. pp In M.P. Zalucki, RAI Drew, GG White eds. Pest Management - Future Challenges. Proceedings of the 6th Australasian Applied Entomology Conference, Brisbane. University of Qld Press, Brisbane. Wilson L.J., Mensah R.K., Fitt, G.P. (2003) Implementing IPM in Australian cotton. Proceeding of the Third World Cotton Conference, Capetown, Sth Africa.

16 Table 1. Scale of field assessment and commercial deployment for INGARD Cotton in Australia Season Area (ha) % of total Main Activity cotton area 1992/ plants - Assessment of outcrossing risk, field efficacy 1993/ Field efficacy, assessment of outcrossing risk 1994/ Field efficacy, environmental impacts (non-targets, pest dynamics) 1995/96 40 (4 sites) - Environmental impacts, IPM performance 1996/97 30,000 8 Five year registration granted, annual review. Limited commercial release by area and region 1997/98 60, Limited commercial release by area and region 1998/99 85, Limited commercial release by area and region 1999/ , Limited commercial release by area and region 2000/ , Limited commercial release by area and region 2001/ , Capped at 30% until two gene Bt varieties available (probably 2003/2004)

17 Table 2. Environmental factors researched as part of Australian pre-release studies. Several other factors were addressed by Monsanto in the USA and accepted by the National Registration Authority. Potential Environmental Risk Gene escape, outcrossing Features and data collection 1. Cultivated cottons tetraploid, all native Gossypium diploid; numerous genetic barriers to successful gene flow 2. Large pollen, not wind dispersed, maximum documented movement only a few metres 3. Cotton essentially selfing species 4. Native Gossypium spp. geographically isolated from cultivated cotton Non- target impacts 1. Large scale biodiversity studies of unsprayed Bt cotton, unsprayed conventional cotton and sprayed conventional cotton fields 2. Approximately 450 species found 3. No detectable impacts on diversity and abundance of most groups of invertebrates, including all groups of generalist predators 4. Significant reduction in specialist parasitoids of Helicoverpa 5. Sprayed conventional cotton had lower abundance of all groups compared to unsprayed Bt cotton Bt cotton as a weed 1. Cotton not regarded as weedy species in current cotton production regions. 2. Bt gene unlikely to confer particular advantage in natural environment Changes in status of minor pests Resistance to Bt protein in target species, Helicoverpa spp. 1. Numerous minor pests occur in Australian cotton (sucking pests, mites) 2. Some of these induced by pesticide use (eg.mites), some suppressed by pesticides applied for Helicoverpa. 3. Latter may become more abundant in Bt cotton with reduced pesticide applications 4. No evidence of such changes in pre-release field studies (1992/ /96) 1. Helicoverpa armigera has history of pesticide resistance 2. Examples of field resistance to Bt in some other species (eg. Diamond back moth, Plutella) 3. Bt resistance genes assumed to occur at frequency of < Comprehensive understanding of H. armigera ecology and behaviour in response to multi-cropping systems 5. Field experiments to identify elements of pre-emptive resistance management strategy, based on refuge crops Conclusion Risk of gene escape effectively zero Llewellyn and Fitt 1996, Brown et al 1997, Brubaker et al 1999 No significant impact on majority of non-target invertebrates found in cotton fields. Minor impacts still considerably less than effects of pesticide sprays on conventional cotton Fitt and Wilson 2002 Potential weediness not regarded as significant risk Experimental studies in northern Australia show no advantage for Bt cottons in most noncropping habitats (Eastick 2002) Short-term, relatively small scale field studies unlikely to detect changed pest dynamics Increased abundance of aphids now occurring after 5 years of commercial use of Bt cotton Resistance is a significant risk. Management strategy developed, implemented from first season of commercial use Strategy components validated and regularly updated Strategy enforced through contracts, legally binding label and compulsory auditing of all farms

18 Table 3. Yield and economic returns from comparative analysis of INGARD and conventional cotton varieties across Australian production regions. Season 1996/ / / / / /02 Crop type Sample size * Licence fee/ha Rebates Net licence fee Yield Bales/ha (Range) INGARD 210 $245 Value guarantee $ ( ) Conventional 8.26 ( ) INGARD 179 $245 $35 $ ( ) Conventional 8.38 ( ) INGARD 110 $185 $30 $ ( ) Conventional 7.39 ( ) INGARD 149 $185 $30 $ ( ) Conventional 7.98 ( ) INGARD 142 $185 $30 $ ( ) Conventional 7.34 ( ) INGARD 229 $185 $15 $ ( ) Conventional 8.77 ( ) Total number of sprays Average Cost /spray Average Insect costs/ha ** (range of values) 5.0 $53 $508 ($410 - $622) 10.3 $45 $467 ($393 - $635) 6.0 $49 $491 ($434 - $515) 10.2 $ ($395 - $524) 8.7 $56 $675 ($418 - $853) 14.0 $52 $766 ($577 - $944) 6.2 $56 $501 ($414 - $656) 10.3 $56 $573 ($205 - $712) 4.6 $57 $426 ($279 - $545) 9.9 $61 $608 ($451 - $704) 3.1 $45 $327 ($260 - $433) 8.6 $53 $504 ($390 - $768) Net Benefit Insect Costs Net Economic Benefit (range) - $41 - $262 (-$409 to +$68) - $35 $22 (-$111 to +$193) $91 $6 (-$1200 to +$1,000) $72 $50 (-$1400 to +$2000) $182 $328 (-$1000 to +$1500) $177 $331 (-$700 to +$1400) * number of paired comparisons of INGARD and conventional cotton crops ** including net INGARD licence fee

19 Table 4. Percentage reduction in spray applications on INGARD cotton varieties throughout crop development. Stage of Crop Development Season Squaring Boll Fill Opening 1997/ / / / /

20 Ingard Conventional Nsprays Helicoverpa Aphids Mirids Mites Thrips Figure 1. Average number of pesticide sprays applied to INGARD and conventional cotton crops in Australia for various pests over the first 6 years of INGARD cotton use. Data based on 120 to 250 paired comparisons each year.

21 Endosulfan Carbamates Synthetic Pyrethroids Organophosphates Figure 2. Average % reduction in number of spray applications of major chemical groups used for Helicoverpa control on Australian cotton (1996/97 to 2001/02).

22 18 16 Number of respondents / / >(1000) (900) to (800) (700) to (600) (500) to (400) (300) to (200) (100) to to to to to to 1000 Economic (Cost)/ Benefit Figure 3. Frequency distribution of economic returns from INGARD cotton relative to conventional cotton over two seasons

23 /97 97/98 98/99 99/00 00/ /02 Figure 4. Net economic benefit ($) of INGARD cotton varieties compared to conventional varieties averaged over all Australian production regions.