Transgenic (Bt) cotton has provided major benefits for the Australia cotton industry (e.g. control of Helicoverpa and reduced pesticide use), but the possibility of Bt resistance in these insects is a major threat. Until recently, Bt resistance research focused only on H. armigera. However, recognition that Bt resistance alleles becoming increasingly common in H. punctigera has moved focus to both species.

As part of a resistance management plan (RMP) for (no exposure to Bt toxins) are required to be grown in association with Bt cotton, at prescribed amounts (ha, as a % of Bt cotton grown), according to historic views of their abilities to produce large numbers of moths that will then disperse, mate with potentially resistant moths emerging from Bt cotton, and thereby reduce the likelihood of Bt resistance emerging to a damaging scale. Initially, several refuge crops were available for use, but this has recently been reduced to only pigeon pea and conventional cotton, because these crops harbour both Helicoverpa species. Pigeon pea is the most commonly used refuge because less (5%) is required cf cotton (10% when unsprayed). However, there is significant desire in industry to move to a less demanding requirement of cotton refuge, because of side benefits it could provide (e.g. cotton yield). Refuge crop options are also less available for dryland than irrigated cotton. In tandem with all this, the amount of Bt cotton on the landscape has drastically changed from a 30% cap of all cotton in the 1990s (Ingard) to 80-90% now (Bollgard II). This background led to this project which sought to answer :

1. How effective are current refuge crop options ?,

2. Are there novel refuge crop options ?,

3. Is the abundance of the target pests, Helicoverpa spp., changing in time at landscape scale ?

The work has shown that there have been substantial changes in the abundance of both Helicoverpa species at landscape scale in recent years. Whilst H. armigera has tracked the abundance of host crops, the numbers of spring migrant (susceptible) H. punctigera have seemingly diminished. The latter has implications for resistance management in this species. Ingard cotton was unable to suppress the abundance of H. armigera, but it is looking increasingly likely that Bollgard II will.

Long-term monitoring of pigeon pea and cotton refuges suggests that the 2 : 1 differential assumed in Helicoverpa productivity has been maintained, but separate work within the project has highlighted the large variability in these refuges’ performance. At times, cotton refuges can be as productive, or more so, than pigeon pea. Many refuges “under-perform”. Modelling is in progress using intensive data sets obtained from surveying refuge crops within regions (St George, Macintyre) to assess the collective performance of refuges in servicing landscapes. Initial results suggest this will demonstrate there are weak links in coverage, where Bt resistance risk will be heightened. Such work will help focus monitoring efforts, and indicate where greater care in placement / management of refuges is needed. This project stresses strongly that high standards of farmer management of refuges (e.g. adequate nutrients, water) are imperative in delivering optimal refuge performance. Perhaps incentives for such management need to be put in place by industry.

Various novel refuge options were explored, especially focusing on improving cotton performance through increased seasonal breadth of attractiveness (e.g. split plantings, mixing varieties and host crop species). However, none of these approaches proved persuasive enough to merit changing options. Pest pressure during the project was generally low, making demonstration of novel refuge options difficult.

A key assumption, implicit in the Bt resistance RMP is that mating occurs at random between moths from different plant host origins. This project supports the assumption. The project also explored the possibility that egg lays (attractiveness) might be correlated with refuge productivity, thereby enabling simpler evaluation of likely refuge performance, in particular relative to Bt cotton. However, such a relationship was not demonstrated clearly. Research in dry- land cotton has been difficult in the past, in part because of the difficulty of traditional methods of assessing refuge productivity (pupae digs, efficiency is compromised in harder soils). This project developed a modified emergence cage which can assist with assessing refuge productivity without the labour of digging.