It is well-recognised that the pest status of Helicoverpa spp. moths within Australian cotton crops has been reduced by the use of Bt cotton varieties, and that much less insecticide is now directed against these pests. But how has the abundance per se of Helicoverpa spp. changed over time within cotton landscapes and how much of the observed change can be attributed to the use of Bt cotton ? Have the two key species, H. armigera and H. punctigera, been influenced to the same extent ? Does the large “spring” immigration of H. punctigera into eastern cotton production landscapes, from the inland where it is unlikely to be influenced by Bt cotton, still occur to the extent it was believed to do in past years ? In part to help answer these questions, CSIRO has been monitoring populations of Helicoverpa at landscape scale in the vicinity of ACRI, Narrabri (Namoi Valley) using pheromone traps since 1992. This work continued in this project.

Irrespective of any numerical suppression of these pests, the possibility of Bt resistance emerging in Helicoverpa spp. remains a major threat for the Australian cotton industry. A Bt Resistance Management Plan (RMP) was therefore established with the introduction of Bt cotton and is regularly reviewed for its effectiveness. Central to the RMP is the use of mandatory refuge crops (currently pigeon pea and conventional cotton) to release large numbers of Bt susceptible moths within cotton production landscapes and thereby limit the development of Bt resistance. At the inception of this project, unexpectedly high frequencies of resistance alleles (especially those for Cry2Ab) had been reported and seemed to be trending upwards, which was of much concern. Our previous research had identified the most productive (although highly variable) refuges. We had also shown that mandatory refuge performance in servicing landscapes with susceptible moths is likely to be very patchy and suggested that such would be inevitably weakened further by the industry decision to halve refuge area requirements. Other work had shown that mating of Helicoverpa moths is random with respect to plant host origin (the latter determined using stable isotope signatures that moths aquire from plants when feeding on them as larvae). Random mating was hitherto a key but untested assumption in the RMP. However, our understanding of the relative contributions of the various plant host origins of moths, be they Bt cotton crops, mandatory refuge crops or other [“unstructured”] refuges, to the total populations of moths throughout landscapes and regions, and the degree of movement / spatial mixing of such moths within and between such areas, is still rudimentary. Such information is critical to an effective RMP. In particular, the scales at which the movements and mixing occur (and their spatial consistency) are directly relevant to the optimal development and use of modelling tools to adequately predict the development of resistance and its management. What is the most appropriate scale to use in such modelling ? Possibly it’s landscape scale, or perhaps something larger is more befitting for very mobile insects ?

In addition, the efficacies of other aspects of the RMP, most notably the mandatory requirement of planting windows for Bt cotton (designed to limit the number of generations of Helicoverpa exposed to Bt toxins) and post-harvest “pupae busting” of Bt cotton (designed to reduce the over-winter survival of resistant Helicoverpa, i.e. from one season to the next), have been questioned. Much has changed in cotton production since the original development of the RMP, or is likely to in the near future (e.g. increased dry-land cotton, greater emphasis on soil health, new tools provided to manipulate crop growth & maturity, the future release of Bollgard 3® cotton with three Bt toxins) which put pressure on the perceived benefits of these RMP-related practices.