Three laboratory strains of Helicoverpa armigera (Hübner) were established by mating of field-collected insects with an existing insecticide-susceptible laboratory strain. These strains were cultured on artificial diet containing the Cry1Ac protoxin of Bacillus thuringiensis using three different protocols. When no response to selection was detected after 7-11 generations of selection, the three strains were combined by controlled mating to preserve genetic diversity. The composite strain (BX) was selected on the basis of growth rate on artificial diet containing Cry1Ac crystals. Resistance to Cry1Ac was first detected after 16 generations of continuous selection. The resistance ratio (RR) peaked approximately 300-fold at generation 21, after which it declined to oscillate between 57- and 111-fold. First-instar H. armigera from generation 25 (RR = 63) were able to complete their larval development on transgenic cotton expressing Cry1Ac and produce fertile adults. There appeared to be a fitness cost associated with resistance on cotton and on artificial diet. The BX strain was not resistant to the commercial Bt spray formulations DiPel and XenTari, which contain multiple insecticidal crystal proteins, but was resistant to the MVP formulation, which only contains Cry1Ac. The strain was also resistant to Cry1Ab but not to Cry2Aa or Cry2Ab. Toxin binding assays showed that the resistant insects lacked the high affinity binding site that was detected in early generations of the strain. Genetic analysis confirmed that resistance in the BX strain of H. armigera is incompletely recessive.

Glasshouse and laboratory experiments were conducted to evaluate the relative fitness of Cry1A-susceptible and laboratory-selected resistant strains of Helicoverpa armigera (Hübner). Life history parameters of H. armigera larvae feeding on young cotton plants showed a significant developmental delay of up to 7 d for the resistant strain compared with the susceptible strain on non-Bacillus thuringiensis (Bt) cotton. This fitness cost was not evident on artificial diet. There was no developmental delay in the F1 hybrid progeny from the reciprocal backcross of the resistant and susceptible strains, indicating that the fitness cost is recessive. In two cohorts tested, survival to pupation of resistant larvae on Bt cotton expressing Cry1Ac was 54 and 51% lower than on non-Bt cotton, whereas all susceptible and F1 larvae tested on Cry1Ac cotton were killed. Mortality of susceptible larvae occurred in the first or second instar, whereas the F1 larvae were able to develop to later instars before dying, demonstrating that resistance is incompletely recessive. The intrinsic rate of increase was reduced by >50% in the resistant strain on Cry1Ac cotton compared with the susceptible strain on non-Bt cotton. There was a significant reduction in the survival of postdiapausal adults from the resistant strain and the F1 strains, indicating that there is a nonrecessive overwintering cost associated with Cry1A resistance in H. armigera.

No abstract