It is evident from experience with the resistance of H. armigera to chemical insecticides and its demonstrated ability to develop resistance to the Cry1Ac toxin produced in INGARD™ that single-gene cotton is not a sustainable control strategy. The second generation of transgenic cotton (BOLLGARD II) prodcues both the Cry1Ac and toxins. This combination was expected to significantly reduce the ease with which H. armigera can develop resistance to these plants. However, a mutagenesis approach adopted by Dr J. Daly has shown that H. armigera is at least capable of to a Cry2A toxin (Cry2a), as well as to Cry1Ac. This result shows that the cotton industry needs to prepare itself for resistance to Cry2Ab as it has with Cry1Ac.

The aim of this project was to determine the magnitude and nature of resistance to Cry2Ab by H. armigera so that an appropriate resistance strategy can be implemented to sustain the use of cry1Ac/cry2Ab transgenic cotton. In the first instance, we set out to establish a Cry2Ab-resistant strain by laboratory selection. Based on our experience in selecting for Cry1A-resistance, we expected that the selection process would be lengthy and that we may only succeed towards the end of the project, at which time we would undertake a preliminary characterisation.

The project was successful beyond expectations in that it produced not one but three strains of H. armigera with resistance to Cry2A. The development of Cry2A resistance in three independent strains after only three generations of selection, coupled with the discovery of Cry2A resistance alleles in field collected insects (in project CSE104C), demonstrates that the optimism that BOLLGARD II™ cotton would provide a solution to the resistance issue may have been misplaced.

Preliminary characterisation of the first of the laboratory-selected Cry2A-resistant strain (TABOC) showed that is not associated with resistance to Cry1Ac and therefore would not be expected to survive on early season BOLLGARD II™. It appears that the resistance detected in the TABOC strain differs from that in the strain obtained by the use of the F2 screen (SP15). It therefore appears that H. armigera has at least two options for resistance to Cry2A.

It is too early at this time to identify the likely consequences of Cry2A resistance in H. armigera. In a successor project to this and CSE104, we will establish whether we have detected one or more types of Cry2A resistance in H. armigera. We shall also determine the dominance of the resistance alleles in the TABOC and SP15 strains and the fitness cost associated with that resistance. This will enable us to develop a refuge strategy to minimise the risk of resistance that would undermine the sustainability of BOLLGARD II™ cotton. Until there is a better understanding of the nature and extent of the threat posed by Cry2A resistance, the cotton industry would be well advised to adopt a cautious attitude to its management of BOLLGARD II™.