In the 1996/97 season the Australian cotton industry adopted an insect-resistant variety of cotton (Ingard®) that is specific to the group of insects including the target Helicoverpa spp. but excluding predators and parasitoids of this pest. To prolong the efficacy of transgenic cotton against Helicoverpa spp., a resistance management plan (RMP) that restricted the area grown to Ingard® was implemented due to the critical importance of preserving the efficacy of the Cry1Ac gene.

In the 2004/5 season Bollgard II® replaced Ingard® as the transgenic variety of cotton available to Australian growers. It improves on Ingard® by incorporating an additional insecticide protein (Cry2Ab) to combat Helicoverpa. Due to the perceived difficulty for Helicoverpa spp. to evolve resistance to both proteins simultaneously within Bollgard II®, the RMP for transgenic cotton was relaxed to allow growers to plant up to 95% of the total area to this product. Bollgard II® was well adopted, with up to 70% (200,000 hectares) planted area throughout the industry.

The sensitivity of field-collected populations of Helicoverpa spp. to Bt products was assayed before and subsequent to the widespread deployment of Ingard® cotton expressing Cry1Ac in the mid-1990’s. From 1994/95 until 2002/03, a Bt spray (MVPII®) that contained formulation ingredients additional to Cry1Ac was used in the screens. The program also incorporated a Bt spray (DiPel®) with insecticidal proteins additional to Cry1Ac to test for resistance to combinations of Cry toxins. The program used only F0 screens however this method cannot detect individuals that are heterozygous for a recessive form of resistance.

During this project we developed screens using a pure Cry1Ac spore/crystal mix as our source of toxin. In anticipation of Bollgard II® replacing Ingard® in 2004/05, we developed methods to screen for resistance to Cry2Ab. In addition to performing F0 screens to detect major changes in gene frequencies, we incorporated an F2 screen to detect and ‘capture’ any rare resistance alleles in natural populations. This method allowed us to simultaneously screen for resistance to Cry1Ac and Cry2Ab, hence making the screens using DiPel® redundant.

There have been no reported field failures of Bollgard II® due to resistance. Our work shows that alleles that confer high level resistance by field populations of H. armigera and H. punctigera are rare for Cry1Ac. However, resistance genes for Cry2Ab in field populations of moths are surprisingly common. Our current best estimate is that they occur for H. armigera at a frequency of 0.004 (upper limit, 0.011; lower limit, 0.0008) and for H. punctigera at a frequency of 0.009 (upper limit, 0.005; lower limit, 0.0001). Individuals that carry a resistance allele for Cry2Ab are killed by Cry1Ac.

Our current knowledge of the ecology and resistance profiles of Australian Helicoverpa populations suggests the RMP is adequate to retard increases in the frequency of resistance. Computer models that incorporate our present knowledge of resistance frequencies, fitness costs, form of dominance and refuge size, suggest that Bollgard II® should prove effective at managing Helicoverpa in the medium to long term. However, it should be emphasised that these models assume that refuges are well maintained in order to produce large numbers of susceptible moths.