Insects can respond to selection pressure by mobilising new defence mechanisms. In contrast to recessive resistance mechanisms based on rare target site mutations in receptor proteins, leading to extreme resistance to high B. thuringiensis (Bt) toxin concentrations, we observed a “hidden” inducible immune and metabolic tolerance in field derived laboratory populations of H. armigera to Cry1Ac and Cry2Ab toxins, with potential to cause ecological and economic problems in cotton production. Given that offspring from insects surviving in the field are tolerant of Bt but do not exhibit resistance to Cry1Ac and Cry2Ab due to target site mutations, it is likely that other mechanism(s) is assisting larvae to survive on certain parts of the plants or late in the season when toxin expression is low. Importantly for cotton bollworm management, we have the following three main outcomes:

1) Tolerance in H. armigera larvae is induced by gut-derived toxins (both Cry1Ac and Cry2Ab) and creates sub-populations of insects that show significant levels of tolerance without displaying mutational changes in a major resistance gene locus. Instead, the tolerant phenotype is caused by differential regulation of immune and metabolic activities (larval induction), and is transmitted to offspring by an epigenetic mechanism showing a maternal effect (embryonic induction). While the epigenetic contribution to incremental increases in tolerance is prominent under low to medium selection pressure, other resistance mechanisms that are transmitted genetically may predominate over time with incremental increase in toxin exposure.

2) Testing gene expression in tolerant and susceptible neonate larvae indicates that a range of genes are expressed significantly differently (both higher and lower expression) in tolerant insects compared to susceptible control. Although some of these differentially expressed genes are important key receptor, putative immune and catalytic genes, most of the differentially regulated genes (up to 190 fold) are unknown in function. By understanding these highly altered unknown genes, it is likely we will gain an understanding of the tolerance mechanism and able to develop management options to specially counter tolerance buildup. Further, significant differential regulation of key receptor genes opens up avenues for further research that may lead to establish monitoring tool to detect inducible tolerance and/or resistance in the field.

3) There appears to be significant developmental penalties under laboratory selection, as evidenced by lowered larval weights and increased developmental times. However, once the populations were kept at constant toxin levels, the developmental penalties slowly diminished over subsequent generations. This could indicate possible genotype selection of allelic combinations of multi-gene functions that reduce developmental penalties. This may in long term provide tolerant populations with the adaptive potential to acquire resistance mechanisms that are genetically transmitted and involve target site mutations in important resistance genes. Our findings have practical significance for adapting pest management protocols to counter inducible tolerance, particularly the importance of susceptible females in blocking this form of resistance.

Lastly, laboratory selected H. armigera (25 generations selected with Cry1Ac, 63 fold resistance) are able to complete their larval development on transgenic cotton expressing Cry1Ac and produced fertile adults (Akhurst et al., 2003). Therefore, whether incremental increases in tolerance can support selection for target site mutations in key receptors allowing insect larvae to survive on Bt-plants is another key issue that requires investigation.