Transgenic cotton has proved to be valued by Australian cotton growers and over 80% of all cotton grown is Bollgard II. The benefits, largely accrued through a reduction in the use of insecticides, have resulted in some cost savings but markedly improved environmental outcomes. A return to a state where 10 sprays per season are required for Helicoverpa control would be unwelcome. Yet the industry is presently reliant on only one transgenic insecticidal variety of cotton - Bollgard II. Resistance to the toxins Cry1Ac and Cry2Ab present in this variety is the greatest challenge to transgenic cotton’s long term sustainability.

Work by CSIRO Entomology has shown that resistance to Cry1Ac remains rare as we have not yet encountered Cry1Ac resistance in the Bt monitoring program. However, we know that an uncommon but potent form of resistance exists in H. armigera populations in China, and it is prudent to assume it is present within Australian populations. Of more immediate concern is the presence of Cry2Ab resistance in Australian populations of this species. We have shown that a variant form (an allele) of a single gene confers resistance to this toxin in H. armigera. That allele is present at a frequency of four in every thousand copies of the gene so it is certainly available to respond to selection.

Since isolating this form of resistance in 2002, CSIRO Entomology’s ‘Bt group’ in Canberra and Narrabri has been studying aspects of the resistance to evaluate the threat it poses to the Australian cotton industry. This project has contributed much to our understanding of the Bt resistance. In addition we possessed a laboratory strain called TABOC, that was selected to be resistant to Cry2Ab, and we also examined its characteristics and compared them to the field-derived form of resistance.

We examined five of the seventeen separate isolations of Cry2Ab resistance in H. armigera to determine their genetic relationship. For all isolations tested to date, resistance was found to result from alleles at the one locus. As the characteristics such as growth and survival rates of larvae were similar for the remaining isolates, we speculate that that they too may be the same form of resistance. In contrast, the resistance present in the laboratory-selected strain TABOC proved to be the result of variants at differing genes or perhaps a constellation of genes.

We examined the performance of a representative (SP15) of the field-derived form of Cry2Ab resistance when fed Bollgard II. Because SP15 is susceptible to Cry1Ac, it performed poorly on younger cotton, but significantly better than susceptible insects on older cotton; presumably when the Cry1Ac titre had declined. Nevertheless, survival rates of resistant insects was low <10%. Importantly, in laboratory tests and when challenged on Bollgard II, larvae carrying one copy of the ‘resistant gene’ (heterozygous) proved to be susceptible to Cry2Ab. This is important as had heterozygotes proved to show an advantage, the threat posed by this form of resistance would have been markedly enhanced.

In many instances where insects develop resistance to an insecticide or to a Bt toxin, individuals carrying the resistance are less fit than susceptible ones. Thus in the absence of the selective agent (in our case host plants other than Cry2Ab-expressing toxin Bollgard II) resistant insects perform poorly. Parameters such as survival and growth rates, fertility and fecundity can be affected in individuals carrying ‘resistant genes’. Such ‘fitness costs’ of resistance tend to retard the evolution of resistance as on non-challenging environments (in our situation, H. armigera growing on refuge crops, weeds or alternative hosts) the frequency of resistance declines. We challenged SP15 under a range of conditions aimed to expose the presence of fitness costs – growth on pigeon pea, conventional cotton during diapause and when exposed to different temperature regimes – however, no costs were detected.

Finally, a component of this project concerned the mechanism of Cry2Ab resistance. An unexpected problem was encountered when this was addressed. The purified Cry2Ab toxin proved to be ‘sticky’ and adhered to cellular material and components of the substrates used to examine its binding. Nevertheless, progress was made in comparing the array of proteins expressed in gut cells that may lead us to identify the causes of resistance.