To inhibit the development of resistance to Bt toxins by Helicoverpa spp, Bt cotton growers in Australia must plant refuges (usually pigeon pea) whose role is to produce large numbers of moths to dilute any resistant genes in moths emerging from Bt cotton. Ideally, Bt cotton should kill all non-resistant larvae. To be effective, refuges need to be attractive to ovipositing moths, and enable larvae to survive to pupation. We compared the attractiveness and productivity of refuges on commercial farms and experimental plots over 5 seasons. We identified factors that increased the attractiveness and productivity of refuges, the genetic makeup of moths emerging from refuges and Bt crops, and the ability of larvae without genetic resistance to survive on Bt cotton.
We found that pigeon pea, was less attractive for oviposition than Bt cotton at the beginning of the season, but tended to be more attractive latter in the season once it started to flower. A higher proportion of moths emerged from Bt cotton than expected. These moths did not carry more resistant genes than those emerging from refuges, but their grandchildren were more tolerant to one of the Bt toxins. Larvae surviving in Bt crops were feeding on flowers or bolls (which contain less Bt toxin). Third instar susceptible larvae in the laboratory developed through to moths on Bt cotton flowers and bolls.
We concluded that refuges are not as attractive or productive as they need to be to effectively dilute resistance genes. Susceptible larvae, especially those moving into the crop as older larvae, survive on Bt cotton by seeking out the structures that produce the least amount of toxin and can pass on their tolerance to their progeny, but how tolerance interacts with genetic resistance is unclear. The continuing efficacy of Bt cotton may indicate the importance of moths entering the system from outside the Bt cotton/refuge matrix.